|
Citation |
- Permanent Link:
- https://ufdc.ufl.edu/UF00084180/00001
Material Information
- Title:
- Statistical modeling and simulation of atmospheric turbulence
- Creator:
- Pahwa, Suresh Badrinath, 1947-
- Publisher:
- University of Florida
- Publication Date:
- 1975
- Language:
- English
- Physical Description:
- xiii, 222 leaves. : illus. ; 28 cm.
Subjects
- Subjects / Keywords:
- Acceleration ( jstor )
Diffusion coefficient ( jstor ) Lagrangian function ( jstor ) Modeling ( jstor ) Prairies ( jstor ) Simulations ( jstor ) Turbulence ( jstor ) Turbulence models ( jstor ) Velocity ( jstor ) Velocity distribution ( jstor ) Atmospheric turbulence ( lcsh ) Chemical Engineering thesis, Ph.D. Dissertations, Academic -- Chemical Engineering -- UF
Notes
- Thesis:
- Thesis -- University of Florida.
- Bibliography:
- Bibliography: leaves 206-221.
- General Note:
- Typescript.
- General Note:
- Vita.
- Statement of Responsibility:
- by Suresh B. Pahwa.
Record Information
- Source Institution:
- University of Florida
- Holding Location:
- University of Florida
- Rights Management:
- Copyright [name of dissertation author]. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
- Resource Identifier:
- 022828021 ( ALEPH )
14101024 ( OCLC ) ADB0691 ( NOTIS )
Aggregation Information
- UFIR:
- Institutional Repository at the University of Florida (IR@UF)
- UFETD:
- University of Florida Theses & Dissertations
- IUF:
- University of Florida
|
Downloads |
This item has the following downloads:
|
Full Text |
Statistical Modeling and Simulation cf
Atmospheric Turbulence
By
SURESH B. PAHWA
A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF
THE UNIVERSITY OF FLORIDA IN PARTIAL
FULFULLMENT OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
1975
ACKNOWLEDGEMENTS
The author wishes to express his sincere appreciation to
Prof. R.W. Fahien, chairman of the supervisory committee, for his
enthusiastic interest and valuable guidance throughout this research.
He is grateful to Dr. D.W. Kirmse, co-chairman of the supervisory
committee, for his interest and many valuable suggestions. He wishes
to thank the other members of his supervisory committee, Prof. R.D.
Walker, Dr. K.E. Gubbins, Dr. A.E.S. Green and Dr. W.C. Huber, for
their cooperation in serving on the committee.
The author wishes to thank Dr. A.W. Westerburg for the use
of the least squares subroutine RMINSQ.
The author gratefully acknowledges the financial assistance
of the College of Engineering, University of Florida. He wishes to
express his gratitude and appreciation to the Department of Chemical
Engineering, University of Florida for both financial support and
personal assistance from its faculty and staff.
Finally, the author extends his thanks to his wife, Brigitte,
whose active support made this work worthwhile.
TABLE OF CONTENTS
Page
ACKNOWLEDGEMENTS................... ..... .............. ii
NOMENCLATURE.......................................... vi
ABSTRACT............... ............. ............... xi
CHAPTER:
1. INTRODUCTION................................... 1
2. PREVIOUS STUDIES ............................... 4
2.1 Structure of Atmospheric Turbulence....... 4
2.1.1 Definition........................... 4
2.1.2 Theories of turbulence................ 5
2.1.3 Turbulence measurements.............. 9
2.1.4 The Monin-Obukhov similarity
theory............................... 12
212 Atmospheric Diffusion...................... 14
2.2.1 Early Developments and the
K-theory............................. 14
2.2.2 Gaussian plume models................. 18
2.2.3 Some recent studies.................. 19
2.2.4 Urban air pollution models........... 21
2.3 Stochastic Models of Turbulence............ 23
3. THEORY............... ...... ........... ....... 27
3.1 Theory of Stochastic Processes............. 27
3.1.1 The Fokker-Planck equation........... 27
3.1.2 Stochastic differential equation..... 33
3.1.3 Comparison of the Fokker-Planck
equation with the diffusion equation. 37
TABLE OF CONTENTS (continued)
3.2 The Growth Law ...........................
3,2.1 The Lagrangian similarity theory
for large times......................
3-2.2 The growth law for a Markov process...
THE STOCHASTIC MODEL............................
4,1 Description of the Model.....................
4.
4.2.2.5
4.2.2.6
The mean time steps...........
The mean Lagrangian
accelerations.................
. RESULTS..........................................
5.1 Computational Parameters....................
Estimation of Simulation Parameters.........
4,2.1 Neutral conditions....................
4.2.1.1 The mean velocity profile.....
4.2.1.2 The variance of the velocities
4.2.1.3 The variance of the
accelerations.................
4.2.1.4 The mean time steps...........
4.2.1.5 The mean Lagrangian
accelerations.................
4,2,2 Non-neutral conditions................
4.2.2.1 The mean velocity profile.....
4.2.2.2 The temperature profile.......
4.2.2.3 The variance of the velocities
4.2.2.4 The variance of the
accelerations................
4,2
Page
40
40
44
52
52
55
59
59
61
TABLE OF CONTENTS (continued)
Page
5.2 Validation of the Model..................... 85
5.2.1 Neutral conditions.................... 86
5.2.2 Non-neutral conditions ................ 96
5.3 Correlated Results.......................... 107
5.4 Concentration Distribution in the
Vertical Direction.......................... 144
5.4.1 Additive function for deviation from
Gaussian behavior...................... 149
5.4.2 Non-Gaussian exponential function..... 153
6. CONCLUSIONS AND RECOMMENDATIONS.................. 159
6.1 Conclusions... ............................. 159
6.2 Recommendations... ......................... 161
APPENDICES:
A. DISCRETIZATION OF DISTRIBUTION CURVES AND
THE MONTE CARLO SIMULATION METHOD............... 163
B. COMPUTER PROGRAMS.................................... 170
LITERATURE CITED....................................... 206
BIOGRAPHICAL SKETCH...................................... 222
NOMENCLATURE
-2
a acceleration, Lt2
a constant in Eq. (4.52)
b constant in Lagrangian similarity theory (Sec. 3.2)
B ratio of the Eulerian and Lagrangian integral length
scales
c concentration, ML3
c concentration distribution coefficient, ML-3
-3
c ground level concentration at y=0, ML
go
2 -2 -1
C specific heat at constant pressure, L t T
2 -1
D molecular mass diffusivity, L t-
E expected value (Sec. 3.1)
E energy spectrum function,L t2
E Gaussian concentration distribution function in the
y y-direction
f deterministic function (Sec. 3.1)
f space correlation coefficient (Sec. 4.2.2.4)
f probability density function (App. A)
F characteristic function of turbulence (Sec. 2.1.4)
2 -1
F energy spectrum function, L t
F concentration in the z-direction
z-
g acceleration due to gravity, Lt-2
g random function (Sec. 3.1)
h deviation function from Gaussian distribution in the
z-direction without reflection
h' reflection term corresponding to h
H stack height, L
H
z
I
k
K
L
L
L
c
LG
ml, mi2
M
n
n
n
N1
N2
Ncell
N
P
p
p
Pl' P2
P
q
q1' q2
Q
- deviation function from Gaussian distribution in the
z-direction
- concentration integral, ML-2
- wave number, L1
2 -1
eddy diffusivity, L T-
- length scale of turbulence, L
- Monin-Obukhov length scale, L
- dimension of length
- length scale for plume spread, L
gradient similarity length scale, L
- exponents in Equations (3.45) and (3.46)
- dimension of mass
-1
- frequency, t-
- partition number for cumulative distribution function
(App. A)
- exponent in Sutton's Lagrangian correlation coeffici-
ent expression (Sec. 2.2.1)
- stability parameter E TLG/6R
- dimensionless pressure gradient
- number of particle samplings in a cell
- total number of particles released
- transition probability
- exponents in Equation (3.40) and (3.41)
-1 -2
- pressure, ML t2
heat flux, Mt3
- constants in Equations (2.14) and (2.15)
source strength, Mt-1.
- source strength, Mt
r generation of the species due to chemical reaction,
ML-3t-1
r exponentially distributed random variable with mean
of one
R gas constant (Sec. 4.2.2.6), MI2t2T-1
R correlation coefficient
s normally distributed random variable with mean=0, and
variance=l
t time
tE Eulerian microscale of turbulence, t
t sampling time, t
T temperature
TG gradient similarity temperature scale, T
T Monin-Obukhov temperature scale, T
T integral time scale of turbulence, t
-l
u velocity in the x-direction, Lt1
S- friction velocity, Lt-1
Up convection velocity, Lt-
v velocity in the y-direction, Lt-
V- velocity vector, Lt1
V c volume of a cell L
w velocity in the z-direction, Lt-
x coordinate axis, downwind distance, L
X displacement of a particle in the x-direction, L
X displacement vector, L
y coordinate axis, cross-wind distance, L
Y displacement of a particle in the y-direction, L
viii
z coordinate axis, vertical distance, L
z roughness length, L
Z displacement of a particle in the z-direction, L
Greek Letters:
a velocity profile exponent
a nt moment of particle displacement, Lnt-l
S- ratio of the Lagrangian and Eulerian integral time
scales
S- constant in the log-linear velocity profile (Sec. 4.2.2)
8 exponent in Eq. (5.60)
Y exponent in Eq. (5.56)
r adiabatic lapse rate, TL-1
6 exponent in F equation (Sec. 5.4)
z
2 -3
e rate of turbulent energy dissipation, L t
n independent variable in the Lagrangian similarity
theory = dt/T
6 potential temperature, T
+ dimensionless potential temperature at the ground
eR
K von Karman constant
A mean value of the time steps
t
A integral length scale of turbulence, L
S- viscosity, ML- t-l
2 -1
v kinematic viscosity, L t-
dimensionless vertical distance (Sec. 5.4.2)
p density, ML-3
a standard deviation
T- lag time, t
T time scale of turbulence (Sec. 3.2), t
-1 -2
S- shear stress (Sec. 4.2.2.1), ML t
-1 -2
T shear stress at the ground, ML t
dimensionless velocity scale
w velocity scale of particle motion, Lt-
Subscripts:
a acceleration
E Eulerian
h heat transfer
i i- direction
L Lagrangian
m momentum transfer
o initial values
o steady state neutral conditions
R reference
t time
v velocity
S' wind direction
Superscripts:
fluctuating quantity
mean quantity
* dimensionless variable
deviation from steady state neutral conditions
t turbulent
Abstract of Dissertation Presented to the Graduate Council
of the University of Florida in Partial Fulfillment of the Requirements
for the Degree of Doctor of Philosophy
STATISTICAL MODELING AND SIMULATION OF
ATMOSPHERIC TURBULENCE
By
Suresh B. Pahwa
March, 1975
Chairman: R.W. Fahien
Co-chairman: D.W. Kirmse
Major Department: Chemical Engineering
A statistical model of turbulent diffusion as applied to flow
in the atmosphere is presented in this work. The model was originally
developed and applied to flow in a pipe by D.W. Kirmse and R.W. Fahien
in 1964.
It is shown that the displacement of a particle as modeled in
this work can be represented as a stochastic process with Markovian
properties. The differential equation for the velocity of the particle
is compared with a generalized stochastic differential equation for a
Markov process. Since the displacement of a particle is assumed to be
Markovian, it obeys a Fokker-Planck equation. Comparison of the Fokker-
Planck equation with the diffusion equation showed that the deterministic
velocity used in the Fokker-Planck equation in general is not equal to
the mean wind velocity, and that the Lagrangian diffusivity is equal to
the Eulerian diffusivity only under certain circumstances.
An approach similar to G.K. Batchelor's similarity theory
was used with the assumption of a Markov process to arrive at a
growth law that expresses the spread of a plume in the vertical
and cross-wind directions, as it moves downwind.
In the simulation, a large number of fluid particles were
released from the source and followed downwind as they move according
to the convection and diffusion in the atmosphere. The particle
displacements are sampled at a fixed interval of time in the downwind
cells to obtain the pollutant concentrations. The particle displace-
ments were simulated by integrating the turbulent Lagrangian accelera-
tions. The total acceleration is the sum of a mean and a fluctuating
acceleration. Estimates of the mean Lagrangian accelerations were
calculated from the equations of motion in the Lagrangian form with
the molecular effects neglected. The initial fluctuating velocities,
the fluctuating accelerations and the time steps were obtained from
the desired probability distribution curves. The statistical parameters
used in the distribution curves were estimated from the turbulent in-
tensities, the mean velocity, the ratio of the Lagrangian to Eulerian
integral time scales, and either the turbulent energy dissipation or
the auto correlation curves.
A sensitivity analysis was carried out to test the effect of
the internal parameters of the model, the total number of particles
released, the sampling time, and the cell size. Optimal values of
these parameters were used in the runs.
The mean Lagrangian accelerations for neutral conditions were
taken as zero. The results obtained from the present work are com-
pared with the experimental atmospheric data of Project Prairie Grass
and with predictions made by other methods.
xii
For non-neutral conditions, the concentration distributions
were found to change with variations in the value of the mean Lagran-
gian acceleration in the vertical direction. However, reasonable
agreement with the experimental data is observed when the mean Lagran-
gian accelerations are taken as zero.
The correlated turbulence data of Cramer and Record for neutral
conditions were used to study the effects of the stack height, the
roughness length and the friction velocity. The results were then cor-
related to obtain generalized charts in dimensionless form for the eddy
diffusivities and the dispersion coefficients as functions of the down-
wind distance with the dimensionless roughness length as the parameter.
xiii
CHAPTER 1
INTRODUCTION
The achievements of the industrial world during the present
century have been accompanied by great ecological disturbances which
have affected the quality of life on the earth. The air we breathe
today is more contaminated than what it was even a decade ago. Eco-
logically balanced urban designs require that the harmful effects of
air pollution be minimal. In the event of a nuclear fallout, one
should be able to predict the spreading of the radio-active material
with sufficient accuracy and speed. Since the Clean Air Act of 1967,
some serious efforts have been made in the abatement of industrial
pollution.
The goals of reaching certain pollution levels are related
to the design of emission control equipment through the use of a dif-
fusion model. A good diffusion model should be able to correlate
pollutant concentration as a function of the rate of emission, source
height, the ground roughness, and meteorological conditions. A sta-
tistical model for turbulent diffusion in the atmosphere from a
steady state, continuous, point source is presented in this work.
Turbulence is stochastic in its behavior. It is not possible
to predict the detailed structure of turbulence at any given time.
Statistical tools enable one to describe turbulence in a comprehen-
sive manner. The stochastic model of turbulence as employed in this
work is capable of using the measured or correlated statistical
properties of turbulence and predicting the diffusion in the atmos-
phere with reasonable accuracy.
The model was originally developed in a Ph.D. thesis by
Kirmse (1964) with Fahien, and applied to turbulent flow in a pipe.
Later Gielow (1972) used the model for flow between parallel plates.
In this work, the model is applied to the atmospheric ground layer.
The model uses a Lagrangian point of view. A Monte Carlo method is
used to generate random values of the Lagrangian accelerations which
are integrated twice to get the fluid particle displacements. A
large number of particles are released from the source and their
paths simulated until they leave the region of interest. The pollu-
tant concentration is obtained by sampling the particle positions in
the cells down wind of the source.
As is shown in Section (3.1), the displacement of the particle
is modeled as a stochastic process with Markovian properties, and thus
it obeys the Fokker-Planck equation which can be considered as a form
of the diffusion equation using the Lagrangian point of view. The
correlation coefficient for a Markov process is of the exponential de-
cay type, and according to Taylor's (1921) theory, an expression for
the standard deviation of the concentration distribution can be derived.
The model was verified with the field observations for differ-
ent stability conditions and reasonably good agreement was observed
with the data. The correlated turbulence data were used to improve
the Pasquill-Gifford charts for the neutral conditions so as to include
3
the effects of the source height and the roughness length. Also,
the eddy diffusivity components were calculated; these can be used
in the solution of the diffusion equation.
CHAPTER 2
PREVIOUS STUDIES
2.1 Structure of Atmospheric Turbulence
2.1.1 Definition
In general, fluids are subjected to two kinds of random
movements: (i) thermal agitation of molecules and (ii) turbulence.
The molecular random movements are discontinuous due to the collisions
of molecules with each other, and with the wall. The molecules have
well defined shapes and sizes which are not dependent upon these
movements. In contrast, turbulence is a continuum motion, and the
eddies change their shapes and sizes continuously. Sutton (1949)
compares the contrast between laminar flow and turbulent flow as
being similar to the contrast between a musical note and a noise in
acoustics.
It is difficult to define turbulence. Perhaps the most
complete definition available in literature is given by Hinze (1959,
page 1): "Turbulent fluid motion is an irregular condition of flow
in which the various quantities show a random variation with time
and space coordinates, so that statistically distinct average values
can be discerned." The researchers in general have found it easier
to describe turbulence than give a formal definition for it.
Bradshaw (1971) describes turbulence as a tangle of vortex lines or
partly rolled vortex sheets in a preferred direction by the mean
flow and in random directions with each other. Corrsin's (1961)
article is a good introduction to turbulence. He describes turbulence
as a random field of vorticity, and states that turbulence can be
expected in a fluid whenever there is a shearing flow and the inertial
effects are much larger than the viscous effects. Lumley and Panofsky
(1964) opted to list the essential features of turbulence. Their
description can be summed up as: Turbulence is a mode of fluid motion
that is a three-dimensional continuum phenomenon, rotational, dissi-
pative, non-linear, stochastic and diffusive with the transport
occurring at time and length scales typically the same as those of
the properties being transferred.
Other introductory discussions of turbulence are given by
Sutton (1953), Schlichting (1955), Townsend (1956), Brodkey (1967),
Batchelor (1967), Phillips (1969), Monin and Yaglom (1971), Tennekes
and Lumley (1972), and Vinnichenko et al. (1973).
2.1.2 Theories of turbulence
Historically, Reynolds (1883) was the first person to visualize
turbulence in a laboratory. His classical experiment consisted of
injecting a dye in a glass tube and observing its mixing behavior.
The smooth flow of the dye with the mean flow changed to rapid lateral
mixing when a critical Reynolds number was reached.
Prior to 1921, turbulence was thought of as consisting of
discrete motions and collisions in an analogous manner to the
molecular motion (Batchelor, 1967). G. I. Taylor (1921) broke with
the idea of analogy to the kinetic theory of gases and introduced
the correlation between the velocities at two points as one of the
quantities needed to describe turbulence. His theory also clearly
shows the enormous simplifications obtained by assuming homogeneity
and isotropy. This assumption has been widely used since to achieve
simplifications in formulation of the turbulence theories.
Richardson (1922) suggested that the energy from the mean
flow is drawn by the large eddies and the large eddies in turn break
up into smaller eddies until the smallest size of eddies, i.e., the
dissipation scale is reached.
Keller and Friedmann in 1924 were the first to state that the
correlation functions and the statistical moments are the fundamental
characteristics of turbulence (Monin and Yaglom, 1971). Using the
equation of motion of a real fluid, Friedmann proposed a general
method of obtaining the differential equations for the statistical
moments of arbitrary order. The total system of equations consists
of an infinite number of equations. Any finite subsystem is always
non-closed. Turbulence modeling techniques associated with the
closure problem are of major interest to investigators in the field
of turbulence.
Based on an analogy to the kinetic theory of gases, Prandtl
(1925) proposed a mixing length of turbulence similar to the mean
free path of the molecules. The theory effectively states that the
turbulent kinematic viscosity is the product of a local characteristic
velocity and the mixing length.
Vt a v'Z
(2.1)
Taylor (1938a) showed the existence of interactions between
components having different length scales. Taylor (1938b) studied the
energy spectrum function of turbulence. The energy spectrum function
describes the distribution of turbulent kinetic energy over various
eddy sizes. He showed that this energy spectrum function is the
Fourier transform of the correlation between the eddy velocities at
two points. Von Karman (1938a and 1938b) showed that the velocity
correlations are tensors. This has enabled more precise and concise
working equations for the non-isotropic situations. Von Karman and
Howarth (1938) reduced these tensors to scalar quantities in the case
of isotropic turbulence. This exemplifies the simplification achieved
through the assumption of isotropy.
Kolmogorov in 1941 put forward the theory of universal
equilibrium. It states that the turbulence energy is generated in
a certain range of length scale and then passed on to smaller and
smaller scales by a cascade process. Thus the sizes are only in-
directly coupled. At high Reynolds numbers, the creation and the
dissipation ranges are widely separated (Ellison, 1962). Obukhov
(1941) arrived at the same conclusions independently. Kolmogorov
arrives at the concept of local isotropy. As the sizes of the
eddies get smaller, there is a tendency for the turbulence structure
to be isotropic. Kolmogorov's theory leads to the conclusion
(Tennekes and Lumley, 1972) that the turbulence energy in the overlap
region of the energy creation range and the Kolmogorov range can be
expressed as
E(k) a 82/3 k-5/3 (2.2)
where E is the energy spectrum function, E is the rate of turbulence
energy dissipation, and k is the wave number. This range is known as
the inertial subrange.
Kolmogorov (1962) and Obukhov (1962) refined their theory of
the small scale structure depending upon one value of the energy
dissipation E (and v) only. In fact, E is a fluctuating quantity,
and the mean E varies with the Reynolds number. A slight deviation
from Eq. (2.2) is observed by taking into account the variations in
the energy dissipation.
Kraichnan (1959) proposed a solution to the closure problem
by expanding the third and higher order statistical moments of the
Fourier transform of the velocity in terms of the second order moments
and their response functions. The response function of the time
correlation coefficient with the lag time T is exp (-vk2 ), where k
is the wave number. He applied the theory to stationary, isotropic
turbulence of very high Reynolds number, where the Reynolds number
was defined in terms of the wave number and the turbulent intensity.
The energy spectrum function in this range was obtained in terms of
a non-linear integral involving the time correlation function and
response functions.
Various closure theories have been put forward by Heisenberg
(1948), Kovasznay (1948), Proudman and Reid (1954), Chandrasekhar
(1955), Tatsumi (1957), Patankar and Spalding (1967), and others.
A survey of developments in the field up to 1970 is discussed by
Reynolds (1974).
2.1.3 Turbulence measurements
One of the more significant sets of experimental measurements
was made by Laufer (1954) for air in fully developed flow in a pipe
at Reynolds numbers from about 50,000 to 500,000. The mean flow and
the statistical properties measurements include the mean velocity
profile, radial, axial and angular turbulence intensities, cross-
correlations between the different components of the fluctuating
velocity, energy spectra and the turbulent kinetic energy dissipa-
tion rate.
In the atmosphere, Richardson (1919,1921,1926) did not use
any of the refined instruments available today, but was able to make
some meaningful approximations and arrive at useful results. Taylor
(1935,1936) studied the behavior of isotropic turbulence in a decaying
wind stream.
Taylor (1938b), using a hot wire anemometer, measured the
kinetic energy associated with different wave numbers, i.e., the
energy spectra. Deissler (1950) studied adiabatic turbulent flows
in pipes. Favre et al. (1953,1957,1958) measured space and time
correlation functions in a wind tunnel and a boundary layer.
Lauffer (1950) conducted a detailed study of the mean and fluctuating
quantities in a two dimensional channel. He made the energy balances
and verified Kolmogorov's hypothesis.
The isotropic turbulence was the object of the studies of
Liepmann and Lauffer (1951), Hanratty (1956) and Corrsin (1958).
In a study of turbulence in a boundary layer over a flat plate
Klebanoff and Diehl (1952) and Klebanoff (1955) found that the
probability densities of the fluctuating velocities are very close
to Gaussian. Other important contributions in the measurements of
laboratory scale turbulence include the works of Rotta (1950),
Liepmann and Robinson (1953), Pai (1953), Sandborn (1955), Sandborn
and Braun (1956), Hanratty (1958), Frenkiel (1958), Johnk and Hanratty
(1962a,1962b), Barbin and Jones (1963), Cermak and Baldwin (1963),
Baldwin and Mickelsen (1963), Clark (1968), Grass (1971), and Snyder
and Lumley (1971).
Sheppard (1946) measured the velocity profile in the atmosphere.
Sheppard used a drag plate at the ground to measure the drag force.
The velocity profile was observed to be logarithmic for near neutral
conditions.
1 f rz+zo
= ( log ] zz (2.3)
where K is the von Karman constant, T is the shear stress on the
ground, p is the density of air, and zo is the roughness length.
The value of the drag coefficient as a function of the roughness
length was evaluated and the variation of the mixing length with
height was studied.
For one observation site, Tait (1949) obtained an empirical
relationship for the temperature gradient in terms of the intensity
of radiation. Lettau (1959) examined the effect of surface roughness
on the geostrophic drag coefficient. The geostrophic drag coefficient
is the ratio of wind speed to the geostrophic speed--where the coriolis
force balances the pressure gradient force. He also discussed the
effect of Richardson number on the shear stress and the velocity
profile. Panofsky and Deland (1959) analyzed the one-dimensional
energy spectra in the lowest one hundred meters of the atmosphere.
They found the Taylor hypothesis regarding equivalence of Eulerian
time and space correlations to be correct, and observed that the
Lagrangian and Eulerian correlations have similar shapes. Record
and Cramer (1966) compared the energy dissipation rates obtained
by previous investigators and found that the variations in c and with
stability and height were extremely large. The turbulence intensity
and the energy spectra data measured at Round Hill were correlated
in the later studies (Cramer et al., 1967; Cramer and Record, 1969).
Other measurements of the turbulent intensities and the energy spectra
include the works of Taylor (1952), Gifford (1955,1956), Panofsky et al.
(1958), Thompson (1962) and Angell (1964).
The study of flux-gradient relations and the temperature
fluctuations in the ground layer have been conducted by Lettau (1950),
Corrsin (1951), Bolgiano (1959,1962), Priestley (1960), Lumley (1964),
Webster (1964), Dyer (1965), Gurvich and Zubkovskii (1966), Clarke
(1970), Okamato and Webb (1970), and Haugen et al. (1971). Haugen
et al. (1971) experimentally found that momentum and heat fluxes are
constant to within 20% provided that a sufficiently long averaging
period is chosen. The velocity-velocity and velocity temperature
correlations were measured and it was shown that the vertical velocity
fluctuation w' is better correlated with temperature than with the
horizontal velocity fluctuation u'.
L
The simulation of the atmospheric ground layer in a wind tunnel
has been of some interest to the experimentalists in recent times.
Strom and Halitsky (1955) outline various factors to be considered in
designing a wind tunnel. Lloyd (1967) generated a shearing flow in
a wind tunnel with an artificially roughened floor. Pruppacher and
Neilburger (1968) discuss the design of a wind tunnel at the University
of California. Arya (1968) and Schon et al. (1974) similated thermally
stratified boundary layers in wind tunnels and compared their observa-
tions with the atmospheric data. Cermak et al. (1966) performed a
dimensional analysis for a wind tunnel and numerous experiments.
2.1.4 The Monin-Obukhov similarity theory
The similarity theory of Monin and Obukhov (1954) by dimensional
analysis arrives at characteristic length and temperatures in a ther-
mally stratified gound layer.
Vertical temperature gradients in the atmosphere create density
differences large enough for the buoyancy term to be included in the
equation of motion. However, based on empirical observations, the
vertical heat and momentum fluxes can be assumed to be constant in
the ground layer. The other assumptions that must be made to derive
the similarity theory are: (i) complete homogeneity exists in the
horizontal (x-y) plane, (ii) variations of density produced by pressure
fluctuations are neglected, (iii) the temperature deviations from the
neutral conditions are small and (iv) the viscous effects are neglected.
Then the three components of the equation of motion, the equation of
continuity, and the thermal energy equation reduce to:
Du 1 8P
Dtu p Px (2.4)
Dt p ax
Dv 1 8P
Dv al (2.5)
Dt p ay
Dw 1 p
Dt p + g T (2.6)
Dt p z TR
Bu av aw
S+ a + 0 (2.7)
ax ay az
where u, v and w are the velocities in the x, y and z directions
respectively, p is the density of air, P and T are the pressure and
temperature deviations respectively from the neutral conditions, g is
the acceleration due to gravity, and TR is the reference temperature.
The basic hypothesis is that the above equations along with
the assumption of the constancy of the heat and momentum fluxes
describe the statistical characteristics of the turbulence in the
ground layer completely. The characteristic parameters for the layer
are then: g/TR, u* and q/C p where u, is the friction velocity. The
only length and temperature scales that can be formed by a combination
of these parameters are:
3
U*
L = (2.8)
T C P
T = (2.9)
Ku, Cp
If the similarity theory holds, the dependence of any mean
turbulence characteristic F on height may be written in the form
F = i (z/L) (2.10)
o
where F is a parameter with the dimensions of F, formed by combination
of g/To, p, u, and q/C p, and t is a universal function of z/L.
Panofsky's (1961) analysis based on the measurements of
Priestley (1960) shows that the similarity theory holds for the
velocity and temperature profiles. The similarity theory was used
by Charnock (1967) to correlate Richardson number, dimensionless
velocity and temperature gradients as functions of the dimensionless
height z/L. The Richardson number can be interpreted as the gravita-
tional force over the inertial force. Taylor (1960), Klug (1968) and
Swinbank (1968) compared the observations made in the atmosphere with
the similarity theory.
Calder (1968) is critical of the similarity theory, and says
that the similarity theory is not a strict consequence of dynamical
equations. There is no closed system of equations for any finite
number of statistical characteristics of turbulence--not even for the
velocity and the temperature gradients. The hypothesis of the existence
of only three parameters, g/TR, u*, q/C p, is a specific viewpoint on
the mechanism of turbulence, and is a hypothesis of maximum simplicity.
2.2 Atmospheric Diffusion
2.2.1 Early developments and the K-theory
The time averaged equation of continuity for a species in a
15
medium can be expressed in general as (Bird et al., 1960):
Dc 2
-+ V-V'c' = DV c + r (2.11)
Dt
where c is the time averaged concentration of the species, V' is the
fluctuating velocity vector, D is the molecular diffusivity, and r is
the rate of production of the species by chemical reaction. A simple
form of Eq. (2.11) is obtained by making the following assumptions:
(i) the turbulent mass flux can be expressed in a manner
analogous to that used by Boussinesq (1877) to describe the turbulent
momentum flux; i.e., it is taken as being proportional to the potential
gradient or
V'c' = K*Vc (2.12)
(ii) molecular diffusion is neglected,
(iii) chemical reaction is neglected.
Roberts (1923) used Eq. (2.12) assuming the eddy diffusivity
to be a scalar quantity (which implies isotropy) that is independent
of position, for the cases of dispersion of a pollutant from (i) an
instantaneous point source, (ii) a continuous point source in a uniform
wind, and (iii) a continuous infinite line source perpendicular to the
direction of a uniform wind. He also obtained an expression for the
visible outline of a smoke plume, and extended the results obtained
with an isotropic diffusivity to anisotropy.
Richardson (1926) considered the diffusion process as the
scatter of a cluster of pollutant particles. He used a scalar dif-
fusivity that was independent of the position in space but was a
function of the effective eddy size by considering the distance
between the diffusing and the neighboring particles. Sutton (1932)
developed the idea of the effective eddy size. He started with
Taylor's (1921) theory
t t'
2 2 f f
2 = 2c f J RL(T) dT dt' (2.13)
0 0
where a2 is the variance of the dispersion, 2 is the variance of the
v
velocity, and RL is the Lagrangian correlation coefficient. He assumed
that
n
RL() = 2- (2.14)
V +a T
n > 0
and derived an expression for a. His result showed that for n < 1, the
size of the effective eddy increases with the distance of the diffusing
cluster from its source.
Sutton (1934) developed an expression for the mass flux due to
evaporation from a surface assuming that the diffusion of mass follows
the same laws as the diffusion of momentum. Sutton (1947) extended his
previous theory of eddy diffusion to include the turbulence intensities
and extended the analysis to heat transfer as well (Sutton, 1948).
Bosanquet and Pearson (1936) proposed a solution of Eq. (2.11)
which is well known as the K-theory. Velocity is assumed independent
of height and the eddy diffusivities in the y and the z-directions
are assumed as follows:
Ky = q u x (2.14)
K = q2 u z (2.15)
where K and K are the eddy mass diffusivity components in the y
y z
and z-directions respectively, q1 and q2 are constants, u is the
constant mean wind velocity, and x and z are the mean wind and the
vertical distances respectively.
Calder (1949) approximated the logarithmic velocity profile
for neutral conditions by the power law profile, and obtained the solu-
tion of the two-dimensional diffusion equation. Calder (1952) made a
contribution to the K-theory by expressing the eddy diffusivity as
K = K u, z (2.16)
z
where u* is the friction velocity, and K is the von Karman constant.
Starting with a power law form of the velocity profile and
assuming the turbulent Schmidt number to be one, Deacon (1949) arrives
at the following form for the vertical turbulent diffusivity:
Kz = K u* z (2.17)
a > 1 unstable conditions
a < 1 stable conditions
Smith (1957) also used a power law form for the vertical eddy
diffusivity and assumed that the eddy diffusivity in the y-
direction also varies in an identical manner with height. He
obtained expressions for ground level concentration profiles from
elevated sources. Further developments of the K-theory include
the works of Davies (1950,1954), Knighting (1952) and Rounds (1955).
2.2.2 Gaussian plume models
Hay and Pasquill (1957) measured dispersion down wind of a
source and showed that the Lagrangian correlations are maintained
high for much longer periods of time than the Eulerian correlations.
This work formed the basis, and later led to the development of the
Gaussian plume models. Hay and Pasquill (1959) explored the idea
further and hypothesized that the Lagrangian and the Eulerian correla-
tions are similar in shapes but have different time scales. Implying
a correlation coefficient equal to one, they assumed
cy Og x (2.18)
where a is the standard deviation of the concentration distribution
y
in the y-direction (cross wind direction), Oa is the standard deviation
of the horizontal wind direction and x is the down wind distance.
However, the correlation coefficient in practice is always less than
one, and Eq. (2.18) should be modified accordingly. Based on the
available data (Haugen, 1959; Monin, 1959; Smith and Hay, 1961; and
a few others), Pasquill (1961) plotted ay vs. x for different 0
values. Similar curves for 0z vs. x were also plotted. In the same
z
paper, Pasquill also showed how to estimate Oa from qualitative
meteorological parameters.
Gifford (1957) showed how similarity predictions of relative
diffusion theory (Batchelor, 1949,1952) apply to the spreading of
smoke puffs, and, using the observations of Kellog (1956) and Frenkiel
and Katz (1956), correlated mean square particle dispersion as a func-
tion of time. Gifford (1959,1961) estimated the diffusion coefficients
in Sutton's (1932) formulation and outlined the method to estimate the
parameters needed to use Pasquill's (1961) graphs. Pasquill and
Gifford both worked independently to estimate the a's and the charts
originally formulated by Pasquill are better known as Pasquill-Gifford
charts. Bowne (1973) has presumably improved the charts by including
the effect of the down wind terrain in a qualitative manner.
The Gaussian plumes are widely used in industry and by the
pollution control agencies (Frenkiel, 1956a; Davidson, 1967; Croke,
1968; Shieh, 1969).
2.2.3 Some recent studies
The number and types of the variables involved in an atmospheric
dispersion problem make the problem a very complex one in nature.
Panofsky (1969) summarizes the different meteorological and physical
parameters involved and their effects. Seinfeld (1969) includes the
effect of chemical reactions and surveys the different diffusion models.
Introductory books on the subject are those of Sutton (1953),
Priestley (1959), Pasquill (1962), Slade (1968), Csanady (1973), and
Seinfeld (1975). The emphasis in the current research work has
shifted to the prediction of dispersion rather than presentation of
theories. This has led to a certain amount of field measurements
and to the estimation of eddy diffusivities.
Stewart et al. (1958) used radioactive argon as a tracer
material from a 61 m stack (BEPO reactor) to study the diffusion of
the plume in the atmosphere down wind of the stack. The effective
values of the constants in Sutton's (1932) diffusivity expressions
were evaluated. Saffman (1962) studied the effect of wind shear on
horizontal spread from an instantaneous ground source. Islitzer and
Dumbauld (1963) showed that the deposition of particulate pollutants
can be significant, thus reducing concentrations at greater distances
to well below that predicted by diffusion models. Fuquay et al.
(1964) (Green Glow program) carried out 46 ground source experiments
and concluded that the standard deviation of the wind direction aa
is not a parameter to use for cross wind diffusion, but the product
u0 should be used instead.
Blackadar (1962) related the turbulent momentum diffusivity
K to wind shear through the use of the expression
m
K = 1/3 4/3 (2.19)
m
where k is a length scale of turbulence. Starting from the turbulent
energy equation, the energy dissipation E was expressed in terms of the
wind shear for neutral conditions. Wu (1965) extended the work for
non-neutral conditions, and, assuming the thermal diffusivity Kh to
be proportional to K derived an expression for Kh.
Hanna (1968) hypothesized that the vertical eddy diffusivity
K depends upon the parameters that determine the energy spectrum
of the vertical velocity fluctuations. The observations of Record
and Cramer (1966), Kaimal (1966), Smith (1967) and others were used
to calculate the momentum, mass and thermal diffusivities. Hosler
(1969) estimated the vertical diffusivities from measurements of
atmospheric radon concentrations.
The modeling of chemical reactions and'incorporating them
into the diffusion equation is discussed by Lamb (1973), Heines and
Peters (1973a,1973b) and Fabrick (1974).
The source height used in the dispersion models is the
effective stack height which is the sum of the physical stack
height and the plume rise. Numerous empirical plume rise models
are available in literature, among which Holland's equation (Thomas
et al., 1963) is the most widely used. Moses and Storm (1961)
compared observed plume rises with some well-known formulas.
Briggs(1965) from dimensional analysis arrived at different equa-
tions for transitional, stable and neutral conditions.
2.2.3 Urban air pollution models
The pioneering work in the field was done by Frenkiel (1956b)
who developed a model for the city of Los Angeles. Lucas (1958) in-
troduced the use of the Gaussian plumes. This approach was adopted
in several modeling studies: Pooler (1961) studied the nature of
atmospheric diffusion over St. Louis, Clark (1964) modeled the city
of Cincinnati, Turner (1964) developed a model for Nashville, and
Miller and Holzworth (1967) for New York City. Koogler et al. (1967)
included a first order reaction rate in Miller and Holzworth's model,
and applied the model to the city of Jacksonville.
Since then various approaches and numerical techniques have
been employed: Panofsky and Prasad (1967), Davidson (1967), Slade
(1967), Sandberg et al. (1970), Eschenroeder and Martinez (1970),
Takamatsu et al. (1970), Johnson et al. (1971), Wayne et al. (1971),
Roth et al. (1971), Lantz et al. (1971), Bowne et al. (1971) and
Aziz et al. (1973). Comprehensive summaries and discussions can be
found in Stern (1970) and Fan and Horie (1971).
The classifications of Moses (1969) and Roth et al. (1971)
are used here to categorize the urban air pollution models.
(i) The source oriented models: The models of this kind use
the integrated form of the diffusion equation and require the
dispersion coefficients as functions of distances from the source,
a complete source inventory and information about other physical
and meteorological parameters.
(ii) The receptor oriented models: The contributions to pollutant
concentrations at a monitoring station are modeled. The sources may
be autos, industry, space heating or power plant stacks. These models
are useful when a large number of sources of different nature are
involved. Several types of receptor oriented models have been developed.
Among these are the Clark model (Clark, 1964), the regression model
(Roberts et al., 1969) and the Argonne tabulation prediction scheme
(Moses, 1969).
(iii) The diffusion equation models: These models are based on
the numerical solution of the diffusion equation including the
chemical reaction term. These models are further divided into:
a) moving cell approach
b) fixed cell approach
In the moving cell approach, concentration changes in a
hypothetical parcel of air are calculated as it is convected with
the wind. The wind velocity is assumed constant with height, and
the diffusion in the horizontal plane is neglected.
In the fixed cell approach, the airshed is divided into a
three-dimensional grid, and the numerical solution may use a finite
difference scheme, a variational method or a particle-in-cell (PIC)
method. The PIC method involves moving the mass particles according
to convection and diffusion.
2.3 Stochastic Models of Turbulence
One of the first stochastic models was developed in a Ph.D.
thesis by Kirmse (1964) and presented by Kirmse and Fahien (1967).
The probability distributions of the random forces acting on the
fluid particles in a random flow were calculated. If the mass of a
fluid particle remains constant, then according to Newton's second
law, the force is proportional to the acceleration. Therefore, it
is sufficient to calculate the probability distributions of the
particle accelerations. The particle's movements are described by
the Lagrangian accelerations as it moves with the flow. The experi-
mental data of Liepmann and Robinson (1953) indicates that the
turbulent velocity and its time derivatives have approximately normal
distributions. This information is used to calculate the statistical
parameters and the paths of the fluid particles are simulated on a
digital computer. This model is used in the present work and is
described in detail in Section 3.1.
Kirmse (1964) simulated the turbulent field in a pipe using
the model, and calculated concentration profiles and eddy diffusivity
components downstream of a source. Gielow (1972) used the same model
for flow between two parallel plates and showed that the model worked
for heat transfer as well. Gielow (1972) also simulated the velocity
record and correlation curves and thus showed that the model was
capable of predicting the detailed structure.
Kirmse's (1964) work must be recognized as a pioneering work
in the numerical simulation of an extremely complicated flow structure
of turbulence. It should also be noted that the Kirmse-Fahien model
is not an exact representation of a turbulent flow field. The
Lagrangian velocity and acceleration distributions are assumed to be
of the same nature. A fluid particle is assumed to maintain its
identity throughout, and the entrapment of mass by the eddies is
neglected. It is also assumed that an equal amount of mass of the
diffusing species is carried by all the eddies. It is however quite
obvious from the results that these effects are small enough to be
neglected in studying the phenomenon of turbulent diffusion.
Davies et al. (1954) and Gupta (1959) noted that Taylor's
(1921) representation of particle diffusion by discontinuous movements
implies zero correlation between the velocities during two different
time steps, and is therefore an inadequate representation of turbulence--
as was also indicated by Taylor (1921).
Kraichnan (1970) simulated an isotropic velocity field in a
turbulent flow. The velocity field was stored in a computer as a set
of Fourier components, which in turn were picked from Gaussian distribu-
tions. Lagrangian velocity correlation, eddy diffusivity and response
functions were calculated and compared with the results obtained by
the direct interaction approximation method (Roberts, 1961).
Deardorff and Peskin's (1970) numerical model involved obtaining the
eddy velocities at each time step from the equation of motion, and
these were integrated to obtain the paths of the particles. The
Lagrangian correlation and mean square dispersion were calculated.
Bullin and Dukler (1974; Bullin, 1972) used a numerical
model similar to the Brownian motion model for the molecular trans-
port, and used a white noise signal to represent the molecular motions.
A major difference between the Kirmse-Fahien and the Bullin-Dukler
stochastic models is the representation of the eddy velocity. An
eddy velocity field continuous in shape is obtained from the
Kirmse-Fahien model (Gielow, 1972) and it possesses the desired
properties. While Bullin and Dukler carefully take into account the
molecular action, they represent the eddy velocity field by a rather
crude approximation of a Gaussian white noise. A white noise consists
entirely of uncorrelated impulses adjacent to each other (Jenkins and
Watts, 1969). Its auto correlation function is a dirac delta function--
which has a zero auto correlation at all times t / 0 and is infinite
at zero time lapse. The impulses are randomly distributed along the
time axis according to a Poisson distribution (Seinfeld and Lapidus,
1974). The model is similar to the Brownian motion model and may
represent molecular transport adequately, but is an oversimplification
of the complicated nature of turbulence.
Goldstein (1951) presented a different analysis of diffusion
by discontinuous movements, although he considered equal time steps
and a velocity of constant magnitude. His model is similar to
Taylor's (1921) analysis of diffusion by discontinuous movements,
with the difference being that Goldstein introduced a two time step
correlation affecting the sign of the velocity. This assumption
results in a telegraph equation in contrast to a first order
differential equation for the one step time dependent process.
However, it is possible to have a non-zero correlation for the one
step dependent process if the velocity and the time steps are not
of constant order of magnitude. Bourret (1960,1961) uses a similar
method to that of Goldstein (1951) but considers the three-dimensional
movement of a particle. Thus he arrives at a sixth order partial
differential equation.
In general, one may expect an equation of infinite order if
the movement of the particle depends on all its previous steps. Such
an analysis is of course inconvenient to deal with. Even the most
complex behaviors in nature, such as the birth and death process and
the genetic effects, can be adequately described by a Markov process
that leads to first order differential equation known as Fokker-Planck
equation.
CHAPTER 3
THEORY
3.1 Theory of Stochastic Processes
A stochastic process is a collection of random variables, or
a chain of events that follow a certain probability law (Tucker,
1967; Kahn, 1956). The random variables can be considered as the
values of a single random variable. A random variable can be defined
as the numerical quantity associated with the outcome of a random
phenomenon (or an experiment) in such a way that, as the various events
or possible outcomes of the phenomenon occur, the random variable
takes on definite values (Parzen, 1962; Feller, 1968). As examples
the random variable could be the instantaneous point value of the con-
centration in a reactor, the instantaneous value of the velocity in a
turbulent flow field, or the displacement of a particle due to turbu-
lent diffusion.
3.1.1 The Fokker-Planck equation
Consider a continuous stochastic process X.. As we have seen
1
above, at any time t, X.(t) can be viewed as the numerical value of
a random variable or we can regard X.(t) as the random variable. For
1
a complete probabilistic description of the stochastic process, all
the moments (infinite in number) are required. Alternatively one
should know all the joint probability density functions; i.e. for any
n, p(Xil,t ; X i,t2; .... X. ,t ) should be known. Using the defi-
i 1 i 2. 1 n
nition of the conditional probability density this can be written as:
p(X ,t; X. t2; ,t ) = p(X ,t2; ....X ,xn X ,tl )p(X ,tl)
1 2 n 2 n 1 1
(3.1)
Since we are interested in p(X. ,t ), the probability density of the
1 n
n
stochastic process X. at time t we can write Eq. (3.1) as:
1 n
p(X ,tl; X 't2 .... ; Xi ,) =
1 2 n
P(Xi ntlXi tn n; ....; X iti)
n n-1 11
P(Xil,t ; X. ,t ); .... X ,t ) (3.2)
1 2 n-1
A Markov process can be defined as a stochastic process where
the present state of the process is completely determined by its pre-
vious state and is independent of the way in which the previous state
had developed. For a Markov process, the conditional probability
density on the right hand side of Eq. (3.2) can be written as:
P(Xi tli ,tn-; Xi 't ; ....; Xit = P(Xin Xi ',tn
n n-1 n-2 1 n n-1
(3.3)
for t1 < t2 < .... t
1 2 n
Eq. (3.3) is the mathematical form of the definition of a
Markov process (Seinfeld and Lapidus, 1974). The conditional prob-
ability p(Xi ,t X. ,tn l) is called the one step transition prob-
n n-1
ability. The probabilistic description of a Markov process does not
require an infinite number of the joint probability density functions
and the problem is reduced to finding the transition probability.
From the definition of a Markov process, it is easy to see that the
transition probabilities obey the Chapman-Kolmogorov equation (Bhar-
ucha-Reid, 1960):
p(X,t Xi ,t ) = fP(Xi,tlX!,t')p(X!,t' IX ,to)dX' (3.4)
o o
-00
where X! = X.(t') is an intermediate state of the stochastic process
1 1
at t', t > t' > t and where X. and t represent the initial values
S1 0
o
of the displacement of the particle and of the time respectively.
The time t' is fixed in the integration of Equation (3.4). The equa-
tion implies that the process can be expressed as going from time t
o
to t in two stages.
Now let us consider an increment of time At. If the present
time in the Chapman-Kolmogorov equation is (t+At), if the intermediate
time is t and if the state of the stochastic process (the particle
displacement is (X.-AX.), Eq. (3.4) can be written as
1 1
00
p(X ,t+AtIX. ,t ) =j P(X.,t+AtlXn-AX.,t)
-00
p(Xi-AX.i,t X ,t )d(AX.)
o
(3.5)
Expanding the integrand on the right hand side of the equation in a
Taylor series in the x-direction about the point p(X.+AX.,t+At|X.,t)-
p(Xi,tJXi ,t ) we obtain:
o
p(X ,t+At iX.-AX.,t)p(X.-AX.,t|X. ,t ) =
0
= p(Xi+AXi,t+Atxi.,t)p(X.tIXi ,to) +
0
(_)n
n! ) n
n
J>1
ax"
i
p(XitlXi ,to)]
o
(3.7)
Substituting Eq. (3.7) into Eq. (3.5), we obtain
S
o
t ) = f p(X.+AXi,t+AtlXi,t)p(Xi,tlX,t )d(AX.)
O 1 1
+ j
-o n=l
(_-)n (A)n
n! i
an
axn
1
[p(X +AX.,it+At Ixi.,t)p.
(Xi,tjXi ,to)]d(AXi)
o
(3.8)
[P(Xi L+!Xi't+AtlXi't)-
P(Xillt+Atlxi
= p(X.,t Xi ,t ) / p(X.+AX.,t+AtlX,t)d(AX.) +
o
--CO
n
[P(Xi,t Xi ,to)
X [ o
l
(AXi)np(X.+AXi,t+AtjX,t)d(AX.)]
1 1 1 1
(3.9)
From the definition of the probability density, we have
0O
f p(X.+AX.,t+At[X.,t)d(AX.) = 1.
Using Equation (3.10), Equation (3.9) can be rewritten as:
Using Equation (3.10), Equation (3.9) can be rewritten as:
p(Xi,t+AtIXi ,to) = p(Xi,tlX ,t ) +
o o
j(AXi)np (Xi+AX, t+At Ii,t)d(AXi )]
-00
n
[p(Xi,tlX. ,to)
i0 o
0
(3.11)
Dividing Equation (3.11) by At and taking the limit as At 0:
Lim
At- 0
p(Xi,t+AtjXi ,t ) -p(X ,tXi ,t ) =
o 1 o
i (-1)n
n!
n=1
an
8 n
X
1
[p(Xi,tli ,t )
0
n
S(AX.)
-00
p(Xi.+AX., t+At ., t)d(AX.)
(3.10)
Lim
At -+ 0
(-_l)n
I n!
n=l
(3.12)
The left hand side of the above equation is simply the definition
of the partial derivative of the transition probability with re-
spect to time.
3p(Xi,tlxi ,t )
0 -
at = Lim
At*O
p(Xi,t+AtIXi ,to)-p(X.,tlXi ,t )
o o
th
The n- moment of a random variable X. with the probability density
function (X) is defined as (Tucker, 1967)
function p(Xi) is defined as (Tucker, 1967)
E {X} = f Xp(X.)d(Xi)
(3.14)
th
Thus the rate of change of the n- moment of the increment AX.
conditioned on (Xt) be defined as
conditioned on (X.,t) can be defined as
1
a. (Xi t) Lim -1 (AX) np(X +AXit+At X.,t)d(AX.)
n 1 At 1 1 1 1 1
1 At0
(3.15)
Substituting Equations (3.13) and (3.15) into equation (3.12), the
"forward Fokker-Planck equation" is obtained.
ap(xi,txi. ,to)
at
n Sai [ac (Xi,t)p(X.,ti x. ,t )]
n! a ni 1 n 0 0
S< X. < t
t > t
(3.13)
(3.16)
The boundary conditions are:
Lim p(X, tI X ,t ) = 6(X.-X )
t-0 o o
(3.17)
Lim P(X.,tlX. ,t ) = 0
Ixl 1 o
Equation (3.16) is also known as the forward Kolmogorov equation.
The three-dimensional Fokker-Planck equation is similar and the de-
rivation is identical to the one described above except for Equation
(3.6) which changes to a three-dimensional Taylor series and has ad-
ditional terms.
3.1.2 Stochastic differential equation
Kirmse (1964) in his Ph.D. thesis describes the displacement
of a particle X. at the end of the time step At as:
im
m
At2
X. = X. + V At + a. m (3.18)
m m-1 m-1 m-1
where X. is the displacement of the particle at the end of the
Im-1
st
(m-l)-- step. V. and a. are the velocity and the acceleration re-
1 1
th
spectively. The displacement of the particle at the end of the m-
step is determined completely from its position at the end of the
previous time step. The single most recent value of the displacement
contains as much information about evaluating X. at t as does the
1 m
m
entire past history of the particle. Therefore the displacement of
a particle in the Kirmse-Fahien model is a Markov process. A sto-
chastic differential equation for the model is derived.in this sec-
tion considering the displacement of a particle as a Markov process.
A general stochastic differential equation for a Markov
process can be written as (Stratonovich, 1968)
dX. dg
S= f (Xit) + (Xi,t) (3.19)
dt i dt
where X. is a stochastic process with Markovian properties, f.(X.,t)
is the deterministic function and g.(X.,t) is the random function of
the process.
The first two moments of the displacement AX. can be calcu-
1
lated as:
a. Li= A 1 f (AX.)p(X+AX.t+Atl.,t)d(AX.)
1. At-mo At f i
1-0
= E-- Xi.t = fi(Xi,t) (3.20)
and
a = (dX) Xt (3.21)
21 dt
The deterministic function f. is then the deterministic
1
th
velocity in the i-- direction in a turbulent flow and the random
function dg./dt is the random velocity contribution in the th
direction. Differentiating Eq. (3.18) with respect to time and
using vector notation, we have
d2X df d2
=-+ (3.22)
dt dt
where X is the displacement vector of a particle in space,
d2X/dt2 is its second derivative with respect to time as it moves
in the space and is therefore the Lagrangian acceleration vector.
The derivative df/dt represents the mean Lagrangian acceleration on
2 2
the particle and (d2g/dt2) is the fluctuating Lagrangian acceleration.
Then
df
a (3.23)
d2
S= a' (3.24)
dt2
where a and a' are the mean and the fluctuating components of the
Lagrangian acceleration vector respectively. Equation (3.22) can
now be written as
2
d2X
= a = a + a' (3.25)
dt2
The Kirmse-Fahien model uses empirical probability distri-
butions for each of the components of the Lagrangian acceleration
and simulates the path of fluid particles on a digital computer.
The model is discussed in detail in Section 4.1.
The integrated form of Eq. (3.25) is the stochastic differ-
ential for a model considering the displacement as a Markov process.
Integrating Eq. (3.25), the stochastic differential equation is ex-
pressed in terms of the Lagrangian accelerations and in the Lagrangian
coordinate system as:
dX t t
d= a dt + V' + a' dt (3.26)
dt = o 0 f
Lag t t
o o
where V' is the turbulent velocity vector possessed by the particle
o
at the beginning of the simulation. The above equation in the Euler-
ian frame of reference is simply the following:
dX t t
= V + a dt + V' + a' dt (3.27)
dt o J
Eul t t
0 0
where V is the mean velocity vector.
3.1.3 Comparison of the Fokker-Planck equation with the diffusion
equation
The Fokker-Planck equation for a = 0 for n > 3 reduces to
n.
i
S1 a2
S(Xti t) = (- p) + 2 (a p) (3.28)
at 0 3X. 1. 2 2 2.
o l l i
1
th
where X. is the particle displacement in the i-- direction. The
1
th
first moment in the i- direction a from Equations (3.20) and
1.
(3.27) then is
t
= fi(Xit) = V. + f a. dt (3.29)
1
t
o
The first moment in the Kirmse-Fahien model a. is therefore
1
the deterministic velocity which is the sum of the average velocity
in the flow field and another term arising due to the mean Lagrangian
acceleration.
The physical meaning of a2. is easy to see from Equation
1
(3.21), which can be alternatively written as
1 dt 2 (3.30)
a = I(d{ X (3.30)
2 2
where dX. is the expected value of (dX.) conditioned at (X.,t).
1 1 1
38
G.I. Taylor (1921) expresses diffusivity as the time derivative of
the variance of dispersion
1 d 2
i 2 dt i (3.31)
Therefore 1/2 a2. is a point eddy mass diffusivity which is
21
I
a function of the particle position and time. The diffusivity 1/2 a2.
is based upon the Lagrangian statistics of the particles (Lamb and
Seinfeld, 1973) and may therefore be called as the Lagrangian diffu-
sivity, KL..
1i
The Fokker-Planck equation can be written in terms of the con-
centration instead of the transition probability. The displacement
X. corresponds to the coordinate x, when we go from probability space
1 1
to time averaged concentration space. Then
p(Xi,t) = c(xi,t) (3.32)
t
al (Xit) = V. (xit) + f a (xi,t) dt (3.33)
t
o
Substituting Equations (3.32) and (3.33) into Equation (3.28) and
using 1/2 a2. = KL we have
1 1
39
t
ac (x ,t) -
('t) = [V.(xi.t) + a (x.,t) dt] c (xit)
t
o
2
+-2 [KL (x.,t) c(x.,t)] (3.34)
8x 1
The time-averaged diffusion equation, including the convec-
tive flow term, is
ac(xi,t)
at 3 x. [Vi(x.,t) c(xi't)]
+ [K (x ,t) c(xi,t)] (3.35)
ax. i i' Dx. 1
1 1
The Eulerian diffusivity K. can be related to the Lagrangian diffu-
1
sivity KL by equating Equation (3.34) and (3.35) and solving for
1
t
c a dt +a (L c)
1 1
t
K. o) (3.36)
1 (ac/Ix)
where t represents the conditions at the source. The two diffu-
sivities are equal to each other if homogeneous conditions exist and
if the Lagrangian accelerations are equal to zero.
The Lagrangian diffusivity in general should be expected to
be a function of the source and of the initial time as well as of
the time of diffusion. A question then arises as to whether the
Eulerian diffusivity K. should also be a function of the source.
1
Molecular diffusivity is a property of the medium and if an eddy dif-
fusivity is to be defined in an analogous manner, it should be a
property of the flow field only rather than a quantity associated
with the source and the time of diffusion.
The rate of change of the moments of the third and higher
order a 's are neglected to obtain correspondence with the diffusion
ni
equation. If the diffusion process has Markovian properties, then
in general the moments of higher orders should be non-zero and should
be retained in the Fokker-Planck equation. The Boussinesq hypothesis
of the flux being proportional to the concentration gradient in that
case is invalid. Several researchers in the field of turbulence have
questioned the basic concept of a turbulent diffusivity, and this
criticism does seem to have some validity. However the diffusivities
are useful empirical coefficients in the flux laws and possess many
advantages in routine dispersion calculations.
3.2 The Growth Law
3.2.1 The Lagrangian similarity theory for large times
The concept of Lagrangian similarity was first introduced by
Batchelor in 1957. If homogeneity exists, a particle moving in the
downstream direction finds itself surrounded by the same turbulent
structure. For free shear, steady state flows, it is assumed that
the structure of the flow is the same at different downstream dis-
tances. Mathematically, this assumption is expressed as:
u(t) u
(t) = F() (3.37)
W(t)
where u(t) is the instantaneous velocity of the particle, u is its
mean velocity, w(t) is a velocity scale of the particle motion at
time t, and the independent variable n is proportional to the ratio
of the time dt to T the time scale of the particle motion.
dt
dr a d (3.38)
T(t)
The function F is stationary and random. The statistical
mean velocity of all the particles is then obtained by averaging Eq.
(3.37) over all the particles. Thus:
dX
dt = u + F(n) w(t) (3.39)
where X is the displacement of the particle. The function F(n) is
obtained by assuming that the length and the velocity scales of the
particle motion are given respectively by:
-_ P-
(3.40)
2,(t) t [X(t)] -
and
W(t) a [X(t)] 2 (3.41)
The time scale of the motion can be obtained by taking the
ratio of the length to the velocity scale.
Pl+P2
T(t) [X(t)] 1 (3.42)
Batchelor (1959) later suggested that one should be able to obtain
the same results for a ground level steady state source in the atmos-
pheric ground layer. He developed this idea in an unpublished work
which is discussed by Ellison (1959), Cermak (1963) and Csanady (1973).
It was assumed that the Eulerian properties of the flow structure de-
pend only upon the friction velocity in a neutrally stratified layer
and that the roughness length affects the mean velocity of the flow
only. The similarity hypothesis for elevated sources is expressed
by Cermak (1963) as follows:
For a marked particle which is at z = H when
t = 0, the statistical properties of particle motion
at time t depend only upon u* and t-tv where t is of
order H/u* or larger, where t is a virtual time ori-
gin with magnitude of order H/u*. (page 51)
The mean longitudinal and the vertical positions of the par-
ticles can be expressed as
dZ
d--t= bu (3.43)
dX t -) (3.
dt z
o
for t>> H
U*
The concentration is assumed to be of the form
m1
c a x (3.
and the plume width at the ground is assumed to be of the form
Y m2 (3.
Y a x (3.2
44)
45)
6)
Based on the two above mentioned relationships, the exponent for
the plume width at the ground is found to be
x
z
m2 log -
z Z
0 0
(3.47)
However, Batchelor's hypothesis (1959) and Cermak's (1963)
developments are for a time of releases larger than H/u*. For a
stack height of 200 m and a friction velocity of 0.4 mps, t should
be larger than 500 seconds. At a velocity of 6 mps, one must be at
and
least 3 km away from the source. Cermak (1963) compared the theory
with some atmospheric diffusion experiments and with some wind tun-
nel data. He concluded that the ratio of source height to roughness
length H/zo is of major importance in atmospheric dispersion calcu-
lations.
3.2.2 The growth law for a Markov process
The Lagrangian similarity theory of Batchelor is modified
for small times based on the correlation coefficient results.
For a Markov process, the correlation coefficient is of the
exponential decay type (Csanady, 1973).
T
RL(T) = e (3.48)
where T is the integral Lagrangian time scale. If Z represents the
displacement of a particle in z-direction, we can write (Taylor, 1921)
2 2 t tt T
22 2 d 2
-- z- = 2 w'2 e dT dt' (3.49)
dt dt2 0
0 0
2
where a is the variance of the dispersion in the z-direction.
z
For neutral conditions (Panofsky and McCormick, 1960; Cramer
and Record, 1969)
w2 u2 (3.50)
w2 2
w' = blu
1 *
where bl is a universal constant.
Let x, y and z be the mean positions of the particles re-
leased from the source. Since the rate of change of z is very small
compared to the rate of change of x, and since axial diffusion can
dx dx
be neglected, one can write -= as the mean velocity of the
dt dt
fluid at z = z (Cermak, 1963). For a logarithmic velocity profile
dx u* z
dt K Z
d~ir11 ^"o
(3.51)
Substituting Equations (3.50) and (3.51) into Eq. (3.49), we have
K* z
( 0 -
d22
dZ 2
2 = 2blu
dx2
X Xt X"
Sf e L dx" dx' (3.52)
0 0
Integrating Eq. (3.52) twice with respect to x, we obtain
the following expression:
2
2 2b K
z = [in --+i- 2
[lnZ +1)]2
z
o
a =bA 1+
z L AL l
Sx -
2 x AL
A 2 A- 1 + e L
L AL l
L
2 3
x x x
A1 +
L2 2!nZ 3!AL3
2 2b K
b 2
In( F + 1)1
where
(3.53)
1
2
(3.54)
(3.55)
Eq. (3.53) can be written as:
A x [ 2 3 1
L x x x 2
z = b 1 + + ... 2
z /TA 3A 12AL A3
L L
(3.56)
Since z changes very little with x, it may be replaced by the stack
height H. Eq. (3.56) can be nondimensionalized by H to obtain an
*
expression for the dispersion coefficient a = az/H:
z
S2 3 -
b x x 2
S 1 ++(3.57)
S 12where
where
/ 2b K
b 1
ln(+ 1
*
z
z -
o H
(3.58)
(3.59)
Expanding Eq. (3.57) we obtain:
*
b x
S=-- x* 1 -
z r 6AL
*2 *3
x x
x x
36AL 270AL
The infinite series above can be approximated by:
(3.60)
47
0 = x (3.61)
z 1
2 1 + x
6AL
or
oz =z (* (3.62)
1 1 + b x
The growth law for a Markov process should therefore be given
by Eq. (3.62). It is important to see if this behavior is obtained
for the empirical correlation coefficient curves as well.
Since the Lagrangian correlation coefficients are assumed to
be similar to the Eulerian correlation coefficients (Sec. 4.2), it is
possible to obtain the shapes of z vs. t curves by integrating Eq.
z
(3.49) twice.
St t'
O a RE(T) dT dt' I (3.63)
S0 0
The experimental data of Prairie Grass was used for the cor-
relation coefficient to obtain the right hand side of Eq. (3.63)
which we will call It. The resulting curves are fitted to the form
t
I+ b (3.64)
t 1 + b2t
This curve fit was carried out for all the stability condi-
tions. The results for one typical neutral (Run #24), one unstable
(Run #61) and one stable (Run #22) conditions are shown in Figures
(3.1) to (3.3). The shape of the growth law obtained from the cor-
relation curves is in excellent agreement with Eq. (3.64) which is
obtained by assuming that the diffusion process is a Markov process
and consequently the correlation curves are of an exponential decay
type.
The same form of the growth law is expected to be valid for
the dispersion in the y-direction as well. Thus
*a bx (3.65)
y H 1 + b x
Y2
Although the effect of stability conditions on the form of
the growth law is assumed negligible, the parameters are functions
of stability conditions, and they should be correlated with the proper
stability parameters.
- / Experimental
t, sec
Figure 3-1. Comparison of It curve calculated from the correlation coefficient data
with the Markov process curve fit for typical neutral conditions.
24
H 16
0
Experimental
8
Markov process curve fit
0 II I I.
0 10 20 30 40 50 60 70
t, sec
Figure 3-2. Comparison of I curve calculated from the correlation coefficient data
with the Markov process curve fit for typical unstable conditions.
1.6
1.2
0.8
0.4
0
t, sec
Figure 3-3.
Comparison of It curve calculated from the correlation coefficient data
with the Markov process curve fit for typical stable conditions.
CHAPTER 4
THE STOCHASTIC MODEL
4.1 Description of the Model
Diffusion from a steady state, continuous point source is
modeled in this work. For the application of the model used in this
work, the origin is at the ground level directly below the source,
the x-axis extends horizontally in the direction of the mean wind,
the y-axis is in the horizontal plane perpendicular to the x-axis
(cross wind), and the z-axis is in the vertical direction. The con-
tinuous emission of pollutants from a stack is assumed to be composed
of a finite large number of mass pockets (of the order of 104). Each
of these mass pockets is referred to as a particle. Each particle
is released from the source with a statistically random velocity.
The motion of the particle in accordance with a stochastic equation
is simulated by a Monte Carlo method until it exits the region of
interest.
Statistical analysis of a velocity field in a steady state
turbulent flow indicates that the velocity has an approximate mar-
ginal Gaussian distribution with constant variance, and that the
time derivatives of the velocities are also continuous and also have
an approximate marginal normal distribution (Simmons and Salter, 1938;
Townsend, 1947). Gaussian distributions with measured or correlated
variances are used to simulate the velocities of the fluid particles.
The variances of the accelerations are calculated from information
regarding the velocity correlations and the relations between the
Lagrangian and Eulerian correlations.
For a digital simulation, it is necessary to discretize the
probability distribution curves (See Appendix A). The particles are
moved in finite time steps, but the simulated instantaneous velocity
field is continuous. In the previous chapter, an equation for the
instantaneous acceleration was derived (Eq. 3.27) from the general
stochastic equation for a Markov process. The displacement AXim of
th th
the particle in the i- direction during the m- time step At is ob-
mt
trained by integrating Eq. (3.25) twice in the i- direction over At .
m
Then
AX, = V At + a. dt + a! dt, (4.1)
m i m i I
At At
m m
where V. is the total velocity of the particle at the beginning
m-1
of the time step. If the Lagrangian. accelerations are assumed con-
stant during the time step At ,
At 2 At 2
m m
AX. = V. At + a + a! (4.2)
1 1 m i 2 1i 22)
m m-1 m-1 m-l
The total velocity V. is the sum of the mean velocity V. and the
i i
m-1 m-1
fluctuating velocity. The model is capable of using any given form
of the mean velocity V. (x,y,z). The mean Lagrangian accelera-
m-1
tions are obtained from the equation of motion by estimating dif-
ferent terms from the measured meteorological parameters. The fluc-
tuating accelerations are obtained randomly from Gaussian distribu-
tions (See Appendix A). The time steps are random and without any
memory, and thus can be obtained from an exponentially distributed
population (Kirmse, 1964). The mean time step can be looked upon as
a time scale of turbulence.
Since it is assumed that no interaction takes place between
individual particles, they are released from the source in succession.
The position of the particle at a fixed interval of time t is stored
s
in the computer to get the concentration values. The concentration
is calculated from the particle sampling according to the following
equation.
Q Ncell(ts)
c= Q cell (4.3)
N V
p cell
where
-1
Q = source strength, MT
N = number of particles released
N = number of particles positions
(samplings) in a cell
t = sampling time, T
V = volume of a cell, M
cell
It can be shown that
Q-- is the source strength represented by each particle.
N
p
Qts
-- is the mass of the pollutant represented by each sampling.
N
p
Qt Ncell
s cel is the mass of the pollutant given by all the samplings
N
p in a cell.
Total reflection is assumed at the ground; i.e., when the
particles hit the ground, they are reflected in a perfectly elastic
manner. The molecular motion may be neglected in the atmosphere be-
cause of very large time and length scales of turbulence (Monin and
Yaglom, 1971; Tennekes and Lumley, 1972).
It is assumed that as soon as a particle is released it
adopts the flow properties of the environment. If a temperature or
a velocity difference exists between the fluid coming out of the
stack and the surrounding fluid, their flow properties will be differ-
ent, and thus the source height should be taken as the effective
stack height and not as the physical height.
4.2 Estimation of Simulation Parameters
A particle is followed as it moves downwind and therefore one
needs the Lagrangian turbulent properties. However because of the
fixed coordinate system used, the mean velocity profile must be known.
For the simulation of Eq. (4.2), the following parameters are
needed: the standard deviations of each of the components of the
fluctuating velocity (turbulence density), the mean Lagrangian ac-
celeration profile, the standard deviations of the fluctuating com-
ponents of the Lagrangian accelerations, and the mean values of the
time steps.
Since only the Eulerian properties are usually measured and
reported, Lagrangian properties are obtained through the use of the
factor
B E A/AL (4.4)
where AE and AL are the Eulerian and Lagrangian macroscopic length
scales respectively. Kirmse (1964) and Gielow (1972) used a value
of.B = 0.7. Hay and Pasquill (1959) propose a constant ratio, g of
the Lagrangian and Eulerian time scales, i.e.
RL(St) = RE(t) (4.5)
where RL and RE are the Lagrangian and Eulerian velocity correla-
tion coefficients respectively. Their estimate of B varied from 1.1
to 8.5 Angell, et al. (1971) reported B from 2.5 to 4.5 However
they estimated B from the ratios of the frequencies at which the
maximas are observed for the Lagrangian and the Eulerian velocity
spectra. These are point values of 8 and do not represent the values
that might be satisfactory over the complete range of frequency
(for spectra) or time (for correlations). Tslitzer and Dumbauld
(1963) estimated B = 5 for all their runs. Since their value is
intermediate between those of Hay and Pasquill, this value is used
throughout this work.
The factor B is calculated from B as follows:
Let TL and TE be the Lagrangian and Eulerian macroscopic
time scales respectively. Then,
TL =J R (Bt) d(Bt) (4.6)
0
and
TE = RE (t) dt (4.7)
0
Expressing TL in terms of TE
TL =fR (st) d(St)
0
j RL (st) dt (4.8)
0
Using Eq. (4.5)
TL = RE (t) dt = E Tg (4.9)
0
= TL/TE
B = AE/AL
EL \\
(4.10)
(4.11)
Using well known relations between the space and time macroscales
(Hinze, 1959)
AL v TL
AE = u TE
and therefore
A uT -
E UTE u
B-=---
A T 6o
V V. v.
i L
th
where a is the intensity of turbulence in the i-- direction.
V.
1
(4.12)
(4.13)
(4.14)
By taking the Fourier transform of Eq. (4.5), the Lagrangian
and Eulerian energy spectra can be correlated as follows (Hay and
Pasquill, 1959):
FL(n) = 6 FE(n)
I
(4.15)
where FL and FE are the Lagrangian and Eulerian energy spectra
respectively, and n is the frequency. Eq. (4.15) when integrated
over the entire frequency range gives the equality between the
Lagrangian and Eulerian turbulence energies. But Bullin and
Dukler (1974) used 0 = 4 in the following equation which is an
approximation of Eq. (4.15).
FL(n) = FE(n) (4.16)
A value of 8 = 4 would imply that the Lagrangian turbulence
energy is four times as large as the Eulerian turbulence energy and
this would lead to serious error in using their model.
The other parameters are calculated as follows:
4.2.1 Neutral conditions: Complete homogeneity is assumed in both
the x and the y directions.
4.2.1.1 The mean velocity profile
The velocity profile for neutral conditions is of the loga-
rithmic form (Monin and Obukhov, 1954). Many derivations are avail-
able in the literature, and the result is independent of the method
of derivation (Lumley and Panofsky, 1964). The derivation presented
here is based on dimensional analysis.
The roughness of an atmospheric ground layer is described
by a roughness length z (also known as "dynamic roughness"). The
roughness length is not equal to the actual height of the vegetation
60
or other physical structures on the ground, but is considerably
smaller in magnitude. It varies from the fraction of a mm over
snow or sand to several meters over cities (Nawrocki and Papa, 1963).
Since the mixing length in the atmosphere is proportional to
(z+zo), the characteristic length scale is taken as (z+z ). The
characteristic velocity in the atmosphere is the friction velocity.
u = (4.17)
where T is the shear stress at the ground and p is the density of
o
the atmosphere. The frictional velocity represents the drag on the
surface. By dimensional analysis, the velocity gradient ( /dz) is
therefore proportional to u /(z+z ). Taking the constant of propor-
0.
1
tionality as -, we have
I u
du 1
(4.18)
dz K (z+z )18)
o
Integrating from z = 0 to z = z, we obtain
u z+zo
u =-- in o (4.19)
K z0
The dimensionless constant K is the von KarmAn constant, and is taken
as 0.4.
The roughness length z is a property of the layer and is
L
61
independent of meteorological conditions. The friction velocity
u* represents the magnitude of the wind velocity and is invariant
with height.
4.2.1.2 The variance of the velocities
These are assumed to be equal to the Eulerian mean square
fluctuations.
o2 = ( ) = (a= 2 (4.20)
v. v. v. E
4.2.1.3 The variance of the accelerations
The Lagrangian acceleration is obtained from the Eulerian
acceleration by the Langrangian-Eulerian transformation using the
ratio of the length scales B (Kirmse, 1964).
-- 2
a =B a ((u)2 (4.21)
a. v. x
ai 1
The standard deviations of the velocity can be obtained either
from the experimental data or from correlated data. The correlation
of the spatial derivative of the velocity can be obtained from the
energy dissipation e if one assumes that small scale structure of
the turbulence is isotropic. This assumption is very reasonable be-
cause isotropy is being assumed only at the dissipation scale of tur-
bulence while the overall behavior is still anisotropic. Then
(Hinze, 1959),
2
E = 15V' ( ) (4.22)
The energy dissipation has been found to be a function of
the wind velocity and height (Record and Cramer, 1966). Record and
Cramer added their Round Hill data (Cramer et al., 1961) to the
composite graph of Ivanov (1962). The dissipation rate and its
height dependence were found to vary significantly with thermal strat-
ification.
In a later report, Cramer and Record (1969) were able to suc-
cessfully correlate all their energy spectra measurements by plotting
F(n)u
all of their data on a graph of dimensionless spectra 2 vs.
u* z
nz
dimensionless frequency This relationship gives the height
and the velocity dependence of the energy spectra, and consequently
the energy dissipation rate. The latter is related to the energy
spectra by the following relation:
0
e = 2v/ k2E(k) dk (4.23)
o
where k is the wave number.
2k n
k = (4.24)
_9
L
E (k) = -- F(n)
2Tr
(4.25)
The wave number k can be non-dimensionalized in a similar manner to
the frequency n if we let
* nz
n -
u
(4.26)
Multiplying Eq. (4.24) by z and substituting Eq. (4.26) into
it, we have
kz nz *
2 -- n
2f d
(4.27)
2
Similarly, dividing Eq. (4.25) by (u*z) and simplifying, gives
27TE u F *
=~ =~ F
2 2
u*z u*z
(4.28)
*
Substituting for E and k in terms of F and n in Eq. (4.23), we
obtain
(4.29)
0o + 2 2
E = 2v J 2n*z ) u z d ( 2
0
and
64
872r 2 2 ,2 ,
E 82 n F dn (4.30)
z
0
2 2
87r2 u* *
2 (4.31)
z
where
2 2
n F dn = (4.32)
0 87r v u
0
The dimensional energy dissipation rate c is taken as a
*
constant and its value obtained from F vs n curves (Cramer and
Record, 1969). Combining Equations (4.21), (4.22) and (4.32), we
obtain an expression for the standard deviation of the Lagrangian
accelerations.
S= 2ru* B /2E* (4.33)
ai z vi
4.2.1.4 The mean time steps
Kirmse (1964) and Gielow (1972) have shown that the mean
value of the time step is proportional to the ratio of the standard
deviations of the velocity and of the acceleration, or
a
v.
X = 0.5 (4.34)
t. a
1 a.
1
1
where the factor 0.5 is determined theoretically.
4.2.1.5 The mean Lagrangian accelerations
The mean Lagrangian acceleration in the x-direction a is
x
the derivative of the turbulent shear stress (Kirmse, 1964; Gielow,
1972). The variation of the shear stress with height is extremely
small in the atmosphere (Monin and Obukhov, 1954; Lumley and Panofsky,
1964) and a is therefore taken as zero. Because of no net buoyancy,
x
the mean Langrangian acceleration in the z-direction is also taken
as zero.
4.2.2 Non-neutral conditions
4.2.2.1 The mean velocity profile
In the case of non-neutral conditions, one must use an addi-
tional length scale called the Monin-Obukhov length L. The expression
for the shear then changes to
= Y ^(4.35)
dz K(Z+Z) (4.35)
where I is a universal function of Z/L. Since L = for neutral con-
z+z
o
editions, = 0 and therefore
L
p (0) = 1
(4.36)
66
The dimensionless shear i can be expanded in a Taylor ser-
ies about zero and for small values of /L (< 0.1), one can
write
z+z \ z+z
-L = 1 + B -- (4.37)
Substituting Eq. (4.37) into Eq. (4.35) and integrating from z = 0
to z = z, we obtain
u = [ in ( z + B (4.38)
K z L
The Monin-Obukhov length is defined as follows in terms of
the turbulent heat flux:
3
U*
L = ) (4.39)
K(g/TR) (q/CpP)
Since the heat flux is difficult to obtain in practice, a convenient
way to avoid this difficulty is by expressing the momentum and heat
fluxes in terms of the velocity and potential temperature gradients.
SKm du (4.40)
pC dz
PC
67
where K and Kh are the momentum and heat turbulent diffusivities
m
respectively. Assuming that the variation in T with height is very
small and can be neglected, we have
T -
o 2 du
-= u = K d
p m dz
(4.42)
Substituting Equations (4.41) and (4.42), Eq. (4.39) becomes
du
uL (Km dz
L=
R
K u, (du dz)
T R idz
(4.43)
(4.44)
A gradient
similarity length LG is defined as
(du
LKh dz
L LL
G Km K dz-
m T R dz
Then the velocity profile equation can be written as
u z+z
u In + G
K z\o) G LG
SG
for zIL| < 0.1
L G
G
(4.45)
(4.46)
where BG is a universal constant. The constant G was taken as
4.5 for unstable conditions and 7.0 for stable conditions (Lumley
and Panofsky, 1964). For IZ/LI > 0.1, Equation (4.37) is not valid,
but a power law equation (Csanady, 1973) can be used:
-- u (zl
u 1 (4.47)
where zl is an arbitrary height. The exponent a depends upon the
roughness length zo, the reference height zl and the stability.
Since the log-linear profile given by Eq. (4.46) can be assumed to
hold up to the height IZ/LG I the reference height
z1 = (4.48)
was chosen as this value. Therefore,the reference velocity is
LG LG
u G+ z
u = L* In( 10 z) +S G 1(4.49)
Sz L
o G
= (zo LG) (4.50)
The gradient similarity length LG describes the stability while zo
describes the physical characteristics. The only dimensionless param-
eters that can be formed from these parameters is Zo/LG. Therefore
1
S= a L 0 (4.51)
The above equation is seen to be obeyed for all the Project
Prairie Grass runs (Barad, 1959). The exponent a is plotted against
the stability parameter z/LG in Fig. (4.1) for all the runs with
stable conditions. The function a can be correlated for stable con-
ditions by the expression
a= a In -) + b (4.52)
Zo
where a is the slope ofa vs. ln( /LG) line and b is the intercept
ZO
at /LG = 1. For unstable conditions
a = a constant
= 0.124 (4.53)
4.2.2.2 The temperature profile
Potential temperature is defined as the temperature a parcel
of air would attain if it was brought to a standard reference pressure
under adiabatic conditions. The potential temperature can be related
to the temperature in the following manner:
Assuming the air to behave as an ideal gas, the first law of
thermodynamics for adiabatic conditions can be written as (Hess, 1954)
RT
C dT dP = 0 (4.54)
p p
0.7
0.6
0.5
0.4
0.3
0.2 -
0 0
Q.1
0 2
10-5 22
10 2
O0
OcD
0
2
zo
ZO/L5
5 102
Figure 4-1. The velocity-profile exponent a vs. the stability parameter z** for stable
conditions.
I r 1
Integrating Eq. (4.54) from the reference pressure PR to P, we have
PR
S= T (-) Cp (4.55)
where 8 is the potential temperature. For the Prairie Grass runs
(Barad, 1959), the pressure profiles were observed to be linear. In
this work, the reference pressure is taken as the pressure at the
ground. Thus the pressure profile is
P = PR + PR g z (4.56)
where pR is the density at the ground, and gz is the acceleration
due to gravity in the z-direction, which is given by
gz = -g = -9.8066 m/sec2 (4.57)
Substituting Eq. (4.56) into Eq. (4.55) and expanding into a bi-
nomial series, we have
PR gZ z R
9 = T(1 ....) (4.58)
PR C
Using the ideal gas law and neglecting the higher order terms, Eq.
(4.58) reduces to
0 = T + -- z (4.59)
C
p
The above equation is a good approximation for Eq. (4.55)
for small values of z.
The potential temperature 6 was plotted against the mean
velocity u (Figs. 4.2 and 4.3) for a large number of Prairie Grass
stable and unstable runs. The slopes of these curves were observed
to be constant. This implies that the gradient similarity length
LG can be taken as a constant in the surface layer (Eq. 4.45). The
potential temperature-mean velocity curves can be expressed as
0 = R + u) (4.60)
Combining Equations (4.46) and (4.60) we have
6 = 6 + ln z + B (4.61)
R du K z GLG
A gradient temperature scale TG analogous to the friction velocity
is defined as
dO
T dz u, (4.62)
G di K
dz
The log-linear temperature profile equation then is
( z+z
e e= R6+T rin 0oo + z (4.63)
R G Lz G LG
10
0 2 4 6 8
u, mps
Figure 4-2.
The potential temperature vs. the.mean velocity
for some Prairie Grass runs with stable conditions.
74
36
#45
#61
32
#5
30
28 -
#16
#152 0
26
24-
22 I I I
0 2 4 6 8
u, mps
Figure 4-3. The potential temperature vs. the mean velocity for
some Prairie Grass runs with unstable conditions.
The gradient temperature scale TG is related to the Monin-Obukhov
temperature T* according to the following equation.
K
T = T
G Kh
(4.64)
Similarly the power law temperature profile is
T i
S= eR + I
R (0.1)t L G
(4.65)
4.2.2.3 The variance of the velocities
4.2.2.3 The variance of the velocities
As in the case of the neutral conditions, the velocity vari-
ances are assumed to be equal to the Eulerian values which may be
obtained from the experimental or correlated data.
4.2.2.4 The variance of the accelerations
The variation of the energy dissipation with stability is not
available. But the functional relationship of with z is assumed
to be the same as for the neutral conditions. This leads to the
0 being inversely proportional to the height. The correlation co-
efficients were measured during the Prairie Grass runs (Haugen, 1959),
and therefore the microscales of turbulence could be estimated. From
Hinze (1959), the time microscale tE. is related to the correlation
1
76
of the velocity derivative according to the following equations:
(4.66)
1 1
2 2
tE
1
f(xi) = RE.
1
(4.67)
where f(x.) is the longitudinal space correlation coefficient. Also
1
ax
2 i_2
vi x.
1
(4.68)
X. =0O
1
According to Taylor's hypothesis
t -u
3t ax
(4.69)
Therefore
(u' x)2
2
a
V.
= --2
--2 2
u tE.
1
Using Equations (4.14) and (4.21), we obtain
r2 a
v.
ai BtE
1 E
(4.70)
(4.71)
4.2.2.5 The mean time steps
The mean time steps are obtained from the turbulent intensity
and the variance of the acceleration in the same manner as for the
case of neutral conditions (Eq. 4.34).
4.2.2.6 The mean .Lagrangian accelerations
The mean accelerations in the x and y directions are taken
as zero. The mean Langrangian acceleration in the z-direction is
obtained from the equation of motion in the z-direction.
Dw 1 3P
a t = t + g (4.72)
z Dt p z z
Let P (z) describe the pressure at steady state neutral conditions.
Then
P = PR + P g z (4.73)
If 3 represents the deviation of the density from po, then
1 1 1 (474)
P + P Po (4.74)
1_ (4.75)
PO po2
0o
78
The deviation p can be calculated by expanding p into a two-dimen-
sional Taylor series about the point po as follows:
T+ P+ (4.76)
P Pp0 T Po
where
T = T To, (4.77)
P = P P (4.78)
and T and P are the mean temperature and pressure profiles re-
o o
spectively for steady state neutral conditions. Assuming the ideal
gas law and neglecting the higher order terms,
P = P + P (4.79)
o o
Substituting Eq. (4.79) into Eq. (4.75), we obtain
1 1 1 T P
po + p (4.80)
P p p2 OT P
1 1 T (4.81)
p p T
0
Substituting Eq. (4.81) into Eq. (4.72), we have
- 1 dP T
a = 1 + + gz
z p dz T P
a0
(4.82)
Since the pressure and temperature profile are known one can there-
fore obtain an expression for a in terms of z. The temperature pro-
z
file for the neutral conditions is given by
T = 68 + r z
o R
(4.83)
where r is the adiabatic lapse rate and is equal to (-g/Cp). Since
the reference pressure is taken as the pressure at the ground level
(4.84)
OR = TR
Substituting Equations (4.59), (4.63), (4.65) and (4.83) into Eq.
(4.77), we have the following expression for T:
( z+z \
S= TG i -zo G LG'^
LG
10
T T
G
S 0-- 1)
(0.1)o
a
z
LG
LG
z> 10
10
(4.85)
(4.86)
1
Combining Equations (4.85) and (4.86) with Eq. (4.83), we have
z+z \
G n --G +
T o G
(4.87)
P
O
C
p
+ + **+
+R+ N z
R R 1
for z < L
10
LZG _
(4.88)
(4.89)
( I + + 1 **)
RoD +~ R1
LG
for z > LG
10
where
* z
z =-
H
z
o
H
** z
z L
LG
T
o
and
T
T
o
(4.90)
(4.91)
(4.92)
8R
TG
(4.93)
(4.94)
FLG
N -
"1-
Similarly, the expression for the pressure deviation term becomes
P + dP
R R dz)
dP
z P(- z
R dz
(4.95)
PR+ (dz)
J dP
P dz dz
S o
A linear approximation for the above equation is
P
P
SdP
z dP o 0]
PR dz dz
R
(4.96)
(4.97)
(4.98)
N N -
** 2 1 R
NC
1+ R- z
R
d -
(4.99)
P-Po
p op
0 0
where
82
Substituting Equations (4.57), (4.98) into Equation (4.82), we
obtain
a 1z R 2 1 R (4.100)
Cp 1 L T + R z*
where /T is given by Equations (4.87) and (4.88). All the terms
in Eq. (4.100) are plotted against the height for a set of typical
conditions in Fig. (4-4). It can be seen that the variation in a
close to the ground is considerable.
close to the ground is considerable.
0.04
0.02
C 1 C **
N N- N z
R 2 2 iR
p S1+N **
0 1 R
-0.02
R N2
1 -"- 2
Cp N
-0.04
0 250 500 750 1000
z, m
Figure 4.4. The variation of different terms in the vertical mean Lagrangian acceleration
equation with height.
CHAPTER 5
RESULTS
5.1 Computational Parameters
The Monte Carlo simulation of the stochastic model used in
this work can be carried out for any arbitrarily chosen number of
particles, sampling time and cell size. A sensitivity analysis was
carried out studying the effect of these parameters, and optimal
values then were used for all the runs. In addition, one is free to
divide the distribution functions for the velocities, accelerations
and time steps intoanynumber of parts (see Appendix A). Gielow
(1972) tried different numbers of partitions in his study and even-
tually used 16 partitions. Since the requirements of precision in
the atmosphere are less than in the laboratory scale simulations,
sixteen partitions were also used in this work for all the distribu-
tion functions.
3 5
The number of particles was varied from 10 to 10 As one
may expect, computer time requirements are roughly proportional to
the number of particles released. The scatter observed in the results
decreases with an increase in the number of particles. However the
marginal utility decreases rapidly with the increase in the number of
particles. From these considerations, 10,000 particles were used in
all the runs.
The sampling time has a similar effect on the results as
does the number of particles. An improvement in the results is ob-
served if the particles are sampled more often; i.e. if the sampling
time is decreased. From a similar analysis, a sampling time of 0.25
seconds was selected.
The available computer storage is an obvious limitation on
the total number of cells. If the downwind distance that the parti-
cles have to travel increases, then the cell size must be increased.
Also since the spread increases approximately linearly with the down-
wind distance, the cell size was varied with the downwind distance;
and instead of a complete network of cells, two-dimensional cell
structures were used at different downwind positions. Since perfect
symmetry is assumed in the y-direction, all the particle positions
are folded in the y-direction. The axis corresponds to the center
of the first cell, and is therefore one half the size of the other
cells. The cell size has no effect on the results except when an ex-
tremely large cell size is used close to the source.
5.2 Validation of the Model
The model was validated by comparing the results of a simu-
lation with those obtained in the Prairie Grass diffusion experiments
(Barad, 1958, 1959; Haugen, 1959). The correlated turbulence data
of Cramer and Record (1969) were used to study the effects of differ-
ent physical and meteorological parameters including the effects of
neglecting the axial diffusion. The results are correlated in a use-
ful form, and graphs for correlated dispersion coefficients and eddy
diffusivities are plotted.
5.2.1 Neutral conditions
Prairie Grass diffusion experiments (Barad, 1958, 1959; Hau-
gen, 1959) were conducted at O'Neill, Nebraska. Sulfur dioxide was
released at a height of 0.46 m from the ground and concentrations
were measured in the horizontal direction along five concentric arcs
and in the vertical direction on six towers. The five arcs were semi-
circular and of radii 50, 100, 200, 400 and 800 meters. All the six
towers were along the 100 m arc. For the same conditions as those of
Run #45 (which was approximately neutral), the concentration distri-
butions in the y-direction at the 50 m, 100 m and 200 m arcs, and in
the z-direction for two towers (y = 5 m and y = 18 m) are compared
in Figures (5-1) to (5-5), respectively, with the experimental re-
sults and with the simulation results of Bullin and Dukler (1974).
The concentration profiles, obtained from the Pasquill-Gifford equa-
tion (Turner, 1967) are also plotted on the same graph using the
standard deviations for the horizontal and vertical dispersion from
the Pasquill-Gifford charts (class D). In all the cases, complete
homogeneity in both the x and the y directions is assumed. It can
be seen that the results are significantly better than using either
of the two above mentioned models.
Run number 24 was seen to be the closest to perfectly neutral
conditions. The results at different downwind distances in the y and
z-directions are plotted in Figs. (5-6) to (5-9). The measured values
of the velocity profile and the turbulent intensities were used in
Near neutral conditions
(LG = -125 m)
x = 50 m
z = 1.5 m
I -
0 Experimental
7
O This work
-- Bullin & Dukler (1974) -
Pasquill-Gifford
(class D)
0 0
- -I O
\
S .0
0/
^Io
\
I 0@
I O
Figure 5-1.
y, m
Concentration distribution in the y-direction for Project Prairie Grass
Run #45 at x = 50.
500
400
300 I-
200 r
to
1 i -.
o\
^
1
|
Full Text |
PAGE 1
6WDWLVWLFDO 0RGHOLQJ DQG 6LPXODWLRQ RI $WPRVSKHULF 7XUEXOHQFH %\ 685(6+ % )$+:$ $ ',66(57$7,21 35(6(17(' 72 7+( *5$'8$7( &281&,/ 2) 7+( 81,9(56,7< 2) )/25,'$ ,1 3$57,$/ )8/)8//0(17 2) 7+( 5(48,5(0(176 )25 7+( '(*5(( 2) '2&725 2) 3+,/2623+< 81,9(56,7< 2) )/25,'$
PAGE 2
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
PAGE 3
7$%/( 2) &217(176 $&.12:/('*(0(176 LL 120(1&/$785( YL $%675$&7 [L &+$37(5 ,1752'8&7,21 35(9,286 678',(6 6WUXFWXUH RI $WPRVSKHULF 7XUEXOHQFH 'HILQLWLRQ 7KHRULHV RI WXUEXOHQFH 7XUEXOHQFH PHDVXUHPHQWV 7KH 0RQLQ2EXNKRY VLPLODULW\ WKHRU\ $WPRVSKHULF 'LIIXVLRQ (DUO\ 'HYHORSPHQWV DQG WKH .WKHRU\ *DXVVLDQ SOXPH PRGHOV 6RPH UHFHQW VWXGLHV 8UEDQ DLU SROOXWLRQ PRGHOV 6WRFKDVWLF 0RGHOV RI 7XUEXOHQFH n 7+(25< 7KHRU\ RI 6WRFKDVWLF 3URFHVVHV 7KH )RNNHU3ODQFN HTXDWLRQ 6WRFKDVWLF GLIIHUHQWLDO HTXDWLRQ &RPSDULVRQ RI WKH )RNNHU3ODQFN HTXDWLRQ ZLWK WKH GLIIXVLRQ HTXDWLRQ
PAGE 4
7$%/( 2) &217(176 FRQWLQXHGf 3DJH 7I¯«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
PAGE 5
7$%/( 2) &217(176 FRQWLQXHGf 3DJH 9DOLGDWLRQ RI WKH 0RGHO 1HXWUDO FRQGLWLRQV 1RQQHXWUDO FRQGLWLRQV &RUUHODWHG 5HVXOWV &RQFHQWUDWLRQ 'LVWULEXWLRQ LQ WKH 9HUWLFDO 'LUHFWLRQ $GGLWLYH IXQFWLRQ IRU GHYLDWLRQ IURP *DXVVLDQ EHKDYLRU 1RQ*DXVVLDQ H[SRQHQWLDO IXQFWLRQ &21&/86,216 $1' 5(&200(1'$7,216 &RQFOXVLRQV 5HFRPPHQGDWLRQV $33(1',&(6 $ ',6&5(7,=$7,21 2) ',675,%87,21 &859(6 $1' 7+( 0217( &$5/2 6,08/$7,21 0(7+2' % &20387(5 352*5$06 /,7(5$785( &,7(' %,2*5$3+,&$/ 6.(7&+ Y
PAGE 6
120(1&/$785( DFFHOHUDWLRQ /W FRQVWDQW LQ (T f FRQVWDQW LQ /DJUDQJLDQ VLPLODULW\ WKHRU\ 6HF f UDWLR RI WKH (XOHULDQ DQG /DJUDQJLDQ LQWHJUDO OHQJWK VFDOHV FRQFHQWUDWLRQ 0/ FRQFHQWUDWLRQ GLVWULEXWLRQ FRHIILFLHQW 0/ JURXQG OHYHO FRQFHQWUDWLRQ DW \ 0/ VSHFLILF KHDW DW FRQVWDQW SUHVVXUH / W 7 PROHFXODU PDVV GLIIXVLYLW\ / W H[SHFWHG YDOXH 6HF f HQHUJ\ VSHFWUXP IXQFWLRQ/ W *DXVVLDQ FRQFHQWUDWLRQ GLVWULEXWLRQ IXQFWLRQ LQ WKH \GLUHFWLRQ GHWHUPLQLVWLF IXQFWLRQ 6HF f VSDFH FRUUHODWLRQ FRHIILFLHQW 6HF f SUREDELOLW\ GHQVLW\ IXQFWLRQ $SS $f FKDUDFWHULVWLF IXQFWLRQ RI WXUEXOHQFH 6HF f 7 HQHUJ\ VSHFWUXP IXQFWLRQ / W FRQFHQWUDWLRQ LQ WKH ]GLUHFWLRQ B DFFHOHUDWLRQ GXH WR JUDYLW\ /W UDQGRP IXQFWLRQ 6HF f GHYLDWLRQ IXQFWLRQ IURP *DXVVLDQ GLVWULEXWLRQ LQ WKH ]GLUHFWLRQ ZLWKRXW UHIOHFWLRQ UHIOHFWLRQ WHUP FRUUHVSRQGLQJ WR ,L VWDFN KHLJKW / YL ,
PAGE 7
+ ; N . e / / / F 0 Q Q GHYLDWLRQ IXQFWLRQ IURP *DXVVLDQ GLVWULEXWLRQ LQ WKH ]GLUHFWLRQ FRQFHQWUDWLRQ LQWHJUDO 0/ ZDYH QXPEHU / HGG\ GLIIXVLYLW\ / 7 OHQJWK VFDOH RI WXUEXOHQFH / 0RQLQ2EXNKRY OHQJWK VFDOH / GLPHQVLRQ RI OHQJWK OHQJWK VFDOH IRU SOXPH VSUHDG / JUDGLHQW VLPLODULW\ OHQJWK VFDOH / H[SRQHQWV LQ (TXDWLRQV f DQG f GLPHQVLRQ RI PDVV IUHTXHQF\ W A SDUWLWLRQ QXPEHU IRU FXPXODWLYH GLVWULEXWLRQ IXQFWLRQ $SS $f H[SRQHQW LQ 6XWWRQf¬V /DJUDQJLDQ FRUUHODWLRQ FRHIILFLn HQW H[SUHVVLRQ 6HF f VWDELOLW\ SDUDPHWHU )/ YD . GLPHQVLRQOHVV SUHVVXUH JUDGLHQW QXPEHU RI SDUWLFOH VDPSOLQJV LQ D FHOO WRWDO QXPEHU RI SDUWLFOHV UHOHDVHG WUDQVLWLRQ SUREDELOLW\ H[SRQHQWV LQ (TXDWLRQ f DQG f SUHVVXUH 0/ W KHDW IOX[ 0W FRQVWDQWV LQ (TXDWLRQV f DQG f VRXUFH VWUHQJWK 0W f¬ YLL
PAGE 8
JHQHUDWLRQ RI WKH VSHFLHV GXH WR FKHPLFDO UHDFWLRQ 0/WB H[SRQHQWLDOO\ GLVWULEXWHG UDQGRP YDULDEOH ZLWK PHDQ RI RQH JDV FRQVWDQW 6HF f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
PAGE 9
] FRRUGLQDWH D[LV YHUWLFDO GLVWDQFH / ] R URXJKQHVV OHQJWK / = GLVSODFHPHQW RI D SDUWLFOH LQ WKH ]GLUHFWLRQ / *UHHN /HWWHUV D YHORFLW\ SURILOH H[SRQHQW D Q Q¯LL PRPHQW RI SDUWLFOH GLVSODFHPHQW /QW UDWLR RI WKH /DJUDQJLDQ DQG (XOHULDQ LQWHJUDO WLPH VFDOHV FRQVWDQW LQ WKH ORJOLQHDU YHORFLW\ SURILOH 6HF f H[SRQHQW LQ (T f \ H[SRQHQW LQ (T f 7 DGLDEDWLF ODSVH UDWH 7/ A H[SRQHQW LQ HTXDWLRQ 6HF f e UDWH RI WXUEXOHQW HQHUJ\ GLVVLSDWLRQ / W Q LQGHSHQGHQW YDULDEOH LQ WKH /DJUDQJLDQ VLPLODULW\ WKHRU\ GWW SRWHQWLDO WHPSHUDWXUH 7 GLPHQVLRQOHVV SRWHQWLDO WHPSHUDWXUH DW WKH JURXQG . YRQ .DUPDQ FRQVWDQW ; W PHDQ YDOXH RI WKH WLPH VWHSV $ LQWHJUDO OHQJWK VFDOH RI WXUEXOHQFH / \ YLVFRVLW\ 0/ W A 9 NLQHPDWLF YLVFRVLW\ / W . GLPHQVLRQOHVV YHUWLFDO GLVWDQFH 6HF f S GHQVLW\ 0/ D VWDQGDUG GHYLDWLRQ L[
PAGE 10
7 ODJ WLPH W 7 WLPH VFDOH RI WXUEXOHQFH 6HF f W 7 VKHDU VWUHVV 6HF f 0/ W 7 2 VKHDU VWUHVV DW WKH JURXQG 0/ W 3 GLPHQVLRQOHVV YHORFLW\ VFDOH ,' f§ YHORFLW\ VFDOH RI SDUWLFOH PRWLRQ /W A 6XEVFULSWV D DFFHOHUDWLRQ ( (XOHULDQ K KHDW WUDQVIHU L Le¯O GLUHFWLRQ / /DJUDQJLDQ P PRPHQWXP WUDQVIHU R LQLWLDO YDOXHV R VWHDG\ VWDWH QHXWUDO FRQGLWLRQV 5 UHIHUHQFH W WLPH 9 YHORFLW\ n ZLQG GLUHFWLRQ 6XSHUVFULSWV I BB IOXFWXDWLQJ TXDQWLW\ PHDQ TXDQWLW\ r GLPHQVLRQOHVV YDULDEOH $ GHYLDWLRQ IURP VWHDG\ VWDWH QHXWUDO FRQGLWLRQV W WXUEXOHQW [
PAGE 11
$EVWUDFW RI 'LVVHUWDWLRQ 3UHVHQWHG WR WKH *UDGXDWH &RXQFLO RI WKH 8QLYHUVLW\ RI )ORULGD LQ 3DUWLDO )XOILOOPHQW RI WKH 5HTXLUHPHQWV IRU WKH 'HJUHH RI 'RFWRU RI 3KLORVRSK\ 67$7,67,&$/ 02'(/,1* $1' 6,08/$7,21 2) $70263+(5,& 785%8/(1&( %\ 6XUHVK % 3DKZD 0DUFK &KDLUPDQ 5: )DKLHQ &RFKDLUPDQ ': .LUPVH 0DMRU 'HSDUWPHQW &KHPLFDO (QJLQHHULQJ $ VWDWLVWLFDO PRGHO RI WXUEXOHQW GLIIXVLRQ DV DSSOLHG WR IORZ LQ WKH DWPRVSKHUH LV SUHVHQWHG LQ WKLV ZRUN 7KH PRGHO ZDV RULJLQDOO\ GHYHORSHG DQG DSSOLHG WR IORZ LQ D SLSH E\ ': .LUPVH DQG 5: )DKLHQ LQ ,W LV VKRZQ WKDW WKH GLVSODFHPHQW RI D SDUWLFOH DV PRGHOHG LQ WKLV ZRUN FDQ EH UHSUHVHQWHG DV D VWRFKDVWLF SURFHVV ZLWK 0DUNRYLDQ SURSHUWLHV 7KH GLIIHUHQWLDO HTXDWLRQ IRU WKH YHORFLW\ RI WKH SDUWLFOH LV FRPSDUHG ZLWK D JHQHUDOL]HG VWRFKDVWLF GLIIHUHQWLDO HTXDWLRQ IRU D 0DUNRY SURFHVV 6LQFH WKH GLVSODFHPHQW RI D SDUWLFOH LV DVVXPHG WR EH 0DUNRYLDQ LW REH\V D )RNNHU3ODQFN HTXDWLRQ &RPSDULVRQ RI WKH )RNNHU 3ODQFN HTXDWLRQ ZLWK WKH GLIIXVLRQ HTXDWLRQ VKRZHG WKDW WKH GHWHUPLQLVWLF YHORFLW\ XVHG LQ WKH )RNNHU3ODQFN HTXDWLRQ LQ JHQHUDO LV QRW HTXDO WR WKH PHDQ ZLQG YHORFLW\ DQG WKDW WKH /DJUDQJLDQ GLIIXVLYLW\ LV HTXDO WR WKH (XOHULDQ GLIIXVLYLW\ RQO\ XQGHU FHUWDLQ FLUFXPVWDQFHV $Q DSSURDFK VLPLODU WR *. %DWFKHORUnV VLPLODULW\ WKHRU\ ZDV XVHG ZLWK WKH DVVXPSWLRQ RI D 0DUNRY SURFHVV WR DUULYH DW D JURZWK ODZ WKDW H[SUHVVHV WKH VSUHDG RI D SOXPH LQ WKH YHUWLFDO [L
PAGE 12
DQG FURVVZLQG GLUHFWLRQV DV LW PRYHV GRZQZLQG ,Q WKH VLPXODWLRQ D ODUJH QXPEHU RI IOXLG SDUWLFOHV ZHUH UHOHDVHG IURP WKH VRXUFH DQG IROORZHG GRZQZLQG DV WKH\ PRYH DFFRUGLQJ WR WKH FRQYHFWLRQ DQG GLIIXVLRQ LQ WKH DWPRVSKHUH 7KH SDUWLFOH GLVSODFHPHQWV DUH VDPSOHG DW D IL[HG LQWHUYDO RI WLPH LQ WKH GRZQZLQG FHOOV WR REWDLQ WKH SROOXWDQW FRQFHQWUDWLRQV 7KH SDUWLFOH GLVSODFHn PHQWV ZHUH VLPXODWHG E\ LQWHJUDWLQJ WKH WXUEXOHQW /DJUDQJLDQ DFFHOHUDn WLRQV 7KH WRWDO DFFHOHUDWLRQ LV WKH VXP RI D PHDQ DQG D IOXFWXDWLQJ DFFHOHUDWLRQ (VWLPDWHV RI WKH PHDQ /DJUDQJLDQ DFFHOHUDWLRQV ZHUH FDOFXODWHG IURP WKH HTXDWLRQV RI PRWLRQ LQ WKH /DJUDQJLDQ IRUP ZLWK WKH PROHFXODU HIIHFWV QHJOHFWHG 7KH LQLWLDO IOXFWXDWLQJ YHORFLWLHV WKH IOXFWXDWLQJ DFFHOHUDWLRQV DQG WKH WLPH VWHSV ZHUH REWDLQHG IURP WKH GHVLUHG SUREDELOLW\ GLVWULEXWLRQ FXUYHV 7KH VWDWLVWLFDO SDUDPHWHUV XVHG LQ WKH GLVWULEXWLRQ FXUYHV ZHUH HVWLPDWHG IURP WKH WXUEXOHQW LQn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n SDUHG ZLWK WKH H[SHULPHQWDO DWPRVSKHULF GDWD RI 3URMHFW 3UDLULH *UDVV DQG ZLWK SUHGLFWLRQV PDGH E\ RWKHU PHWKRGV [LL
PAGE 13
)RU QRQQHXWUDO FRQGLWLRQV WKH FRQFHQWUDWLRQ GLVWULEXWLRQV ZHUH IRXQG WR FKDQJH ZLWK YDULDWLRQV LQ WKH YDOXH RI WKH PHDQ /DJUDQ JLDQ DFFHOHUDWLRQ LQ WKH YHUWLFDO GLUHFWLRQ +RZHYHU UHDVRQDEOH DJUHHPHQW ZLWK WKH H[SHULPHQWDO GDWD LV REVHUYHG ZKHQ WKH PHDQ /DJUDQ JLDQ DFFHOHUDWLRQV DUH WDNHQ DV ]HUR 7KH FRUUHODWHG WXUEXOHQFH GDWD RI &UDPHU DQG 5HFRUG IRU QHXWUDO FRQGLWLRQV ZHUH XVHG WR VWXG\ WKH HIIHFWV RI WKH VWDFN KHLJKW WKH URXJKQHVV OHQJWK DQG WKH IULFWLRQ YHORFLW\ 7KH UHVXOWV ZHUH WKHQ FRUn UHODWHG WR REWDLQ JHQHUDOL]HG FKDUWV LQ GLPHQVLRQOHVV IRUP IRU WKH HGG\ GLIIXVLYLWLHV DQG WKH GLVSHUVLRQ FRHIILFLHQWV DV IXQFWLRQV RI WKH GRZQn ZLQG GLVWDQFH ZLWK WKH GLPHQVLRQOHVV URXJKQHVV OHQJWK DV WKH SDUDPHWHU [LLL
PAGE 14
&+$37(5 ,1752'8&7,21 7KH DFKLHYHPHQWV RI WKH LQGXVWULDO ZRUOG GXULQJ WKH SUHVHQW FHQWXU\ KDYH EHHQ DFFRPSDQLHG E\ JUHDW HFRORJLFDO GLVWXUEDQFHV ZKLFK KDYH DIIHFWHG WKH TXDOLW\ RI OLIH RQ WKH HDUWK 7KH DLU ZH EUHDWKH WRGD\ LV PRUH FRQWDPLQDWHG WKDQ ZKDW LW ZDV HYHQ D GHFDGH DJR (FRn ORJLFDOO\ EDODQFHG XUEDQ GHVLJQV UHTXLUH WKDW WKH KDUPIXO HIIHFWV RI DLU SROOXWLRQ EH PLQLPDO ,Q WKH HYHQW RI D QXFOHDU IDOORXW RQH VKRXOG EH DEOH WR SUHGLFW WKH VSUHDGLQJ RI WKH UDGLRDFWLYH PDWHULDO ZLWK VXIILFLHQW DFFXUDF\ DQG VSHHG 6LQFH WKH &OHDQ $LU $FW RI VRPH VHULRXV HIIRUWV KDYH EHHQ PDGH LQ WKH DEDWHPHQW RI LQGXVWULDO SROOXWLRQ 7KH JRDOV RI UHDFKLQJ FHUWDLQ SROOXWLRQ OHYHOV DUH UHODWHG WR WKH GHVLJQ RI HPLVVLRQ FRQWURO HTXLSPHQW WKURXJK WKH XVH RI D GLIn IXVLRQ PRGHO $ JRRG GLIIXVLRQ PRGHO VKRXOG EH DEOH WR FRUUHODWH SROOXWDQW FRQFHQWUDWLRQ DV D IXQFWLRQ RI WKH UDWH RI HPLVVLRQ VRXUFH KHLJKW WKH JURXQG URXJKQHVV DQG PHWHRURORJLFDO FRQGLWLRQV $ VWDn WLVWLFDO PRGHO IRU WXUEXOHQW GLIIXVLRQ LQ WKH DWPRVSKHUH IURP D VWHDG\ VWDWH FRQWLQXRXV SRLQW VRXUFH LV SUHVHQWHG LQ WKLV ZRUN 7XUEXOHQFH LV VWRFKDVWLF LQ LWV EHKDYLRU ,W LV QRW SRVVLEOH WR SUHGLFW WKH GHWDLOHG VWUXFWXUH RI WXUEXOHQFH DW DQ\ JLYHQ WLPH 6WDWLVWLFDO WRROV HQDEOH RQH WR GHVFULEH WXUEXOHQFH LQ D FRPSUHKHQn VLYH PDQQHU 7KH VWRFKDVWLF PRGHO RI WXUEXOHQFH DV HPSOR\HG LQ WKLV
PAGE 15
ZRUN LV FDSDEOH RI XVLQJ WKH PHDVXUHG RU FRUUHODWHG VWDWLVWLFDO SURSHUWLHV RI WXUEXOHQFH DQG SUHGLFWLQJ WKH GLIIXVLRQ LQ WKH DWPRVn SKHUH ZLWK UHDVRQDEOH DFFXUDF\ 7KH PRGHO ZDV RULJLQDOO\ GHYHORSHG LQ D 3K' WKHVLV E\ .LUPVH f ZLWK )DKLHQ DQG DSSOLHG WR WXUEXOHQW IORZ LQ D SLSH /DWHU *LHORZ f XVHG WKH PRGHO IRU IORZ EHWZHHQ SDUDOOHO SODWHV ,Q WKLV ZRUN WKH PRGHO LV DSSOLHG WR WKH DWPRVSKHULF JURXQG OD\HU 7KH PRGHO XVHV D /DJUDQJLDQ SRLQW RI YLHZ $ 0RQWH &DUOR PHWKRG LV XVHG WR JHQHUDWH UDQGRP YDOXHV RI WKH /DJUDQJLDQ DFFHOHUDWLRQV ZKLFK DUH LQWHJUDWHG WZLFH WR JHW WKH IOXLG SDUWLFOH GLVSODFHPHQWV $ ODUJH QXPEHU RI SDUWLFOHV DUH UHOHDVHG IURP WKH VRXUFH DQG WKHLU SDWKV VLPXODWHG XQWLO WKH\ OHDYH WKH UHJLRQ RI LQWHUHVW 7KH SROOXn WDQW FRQFHQWUDWLRQ LV REWDLQHG E\ VDPSOLQJ WKH SDUWLFOH SRVLWLRQV LQ WKH FHOOV GRZQ ZLQG RI WKH VRXUFH $V LV VKRZQ LQ 6HFWLRQ f WKH GLVSODFHPHQW RI WKH SDUWLFOH LV PRGHOHG DV D VWRFKDVWLF SURFHVV ZLWK 0DUNRYLDQ SURSHUWLHV DQG WKXV LW REH\V WKH )RNNHU3ODQFN HTXDWLRQ ZKLFK FDQ EH FRQVLGHUHG DV D IRUP RI WKH GLIIXVLRQ HTXDWLRQ XVLQJ WKH /DJUDQJLDQ SRLQW RI YLHZ 7KH FRUUHODWLRQ FRHIILFLHQW IRU D 0DUNRY SURFHVV LV RI WKH H[SRQHQWLDO GHn FD\ W\SH DQG DFFRUGLQJ WR 7D\ORUnV f WKHRU\ DQ H[SUHVVLRQ IRU WKH VWDQGDUG GHYLDWLRQ RI WKH FRQFHQWUDWLRQ GLVWULEXWLRQ FDQ EH GHULYHG 7KH PRGHO ZDV YHULILHG ZLWK WKH ILHOG REVHUYDWLRQV IRU GLIIHUn HQW VWDELOLW\ FRQGLWLRQV DQG UHDVRQDEO\ JRRG DJUHHPHQW ZDV REVHUYHG ZLWK WKH GDWD 7KH FRUUHODWHG WXUEXOHQFH GDWD ZHUH XVHG WR LPSURYH WKH 3DVTXLOO*LIIRUG FKDUWV IRU WKH QHXWUDO FRQGLWLRQV VR DV WR LQFOXGH
PAGE 16
WKH HIIHFWV RI WKH VRXUFH KHLJKW DQG WKH URXJKQHVV OHQJWK $OVR WKH HGG\ GLIIXVLYLW\ FRPSRQHQWV ZHUH FDOFXODWHG WKHVH FDQ EH XVHG LQ WKH VROXWLRQ RI WKH GLIIXVLRQ HTXDWLRQ ? ?
PAGE 17
&+$37(5 35(9,286 678',(6 6WUXFWXUH RI $WPRVSKHULF 7XUEXOHQFH 'HILQLWLRQ ,Q JHQHUDO IOXLGV DUH VXEMHFWHG WR WZR NLQGV RI UDQGRP PRYHPHQWV Lf WKHUPDO DJLWDWLRQ RI PROHFXOHV DQG LLf WXUEXOHQFH 7KH PROHFXODU UDQGRP PRYHPHQWV DUH GLVFRQWLQXRXV GXH WR WKH FROOLVLRQV RI PROHFXOHV ZLWK HDFK RWKHU DQG ZLWK WKH ZDOO 7KH PROHFXOHV KDYH ZHOO GHILQHG VKDSHV DQG VL]HV ZKLFK DUH QRW GHSHQGHQW XSRQ WKHVH PRYHPHQWV ,Q FRQWUDVW WXUEXOHQFH LV D FRQWLQXXP PRWLRQ DQG WKH HGGLHV FKDQJH WKHLU VKDSHV DQG VL]HV FRQWLQXRXVO\ 6XWWRQ f FRPSDUHV WKH FRQWUDVW EHWZHHQ ODPLQDU IORZ DQG WXUEXOHQW IORZ DV EHLQJ VLPLODU WR WKH FRQWUDVW EHWZHHQ D PXVLFDO QRWH DQG D QRLVH LQ DFRXVWLFV ,W LV GLIILFXOW WR GHILQH WXUEXOHQFH 3HUKDSV WKH PRVW FRPSOHWH GHILQLWLRQ DYDLODEOH LQ OLWHUDWXUH LV JLYHQ E\ +LQ]H SDJH f 7XUEXOHQW IOXLG PRWLRQ LV DQ LUUHJXODU FRQGLWLRQ RI IORZ LQ ZKLFK WKH YDULRXV TXDQWLWLHV VKRZ D UDQGRP YDULDWLRQ ZLWK WLPH DQG VSDFH FRRUGLQDWHV VR WKDW VWDWLVWLFDOO\ GLVWLQFW DYHUDJH YDOXHV FDQ EH GLVFHUQHG 7KH UHVHDUFKHUV LQ JHQHUDO KDYH IRXQG LW HDVLHU WR GHVFULEH WXUEXOHQFH WKDQ JLYH D IRUPDO GHILQLWLRQ IRU LW %UDGVKDZ f GHVFULEHV WXUEXOHQFH DV D WDQJOH RI YRUWH[ OLQHV RU SDUWO\ UROOHG YRUWH[ VKHHWV LQ D SUHIHUUHG GLUHFWLRQ E\ WKH PHDQ
PAGE 18
IORZ DQG LQ UDQGRP GLUHFWLRQV ZLWK HDFK RWKHU &RUUVLQnV f DUWLFOH LV D JRRG LQWURGXFWLRQ WR WXUEXOHQFH +H GHVFULEHV WXUEXOHQFH DV D UDQGRP ILHOG RI YRUWLFLW\ DQG VWDWHV WKDW WXUEXOHQFH FDQ EH H[SHFWHG LQ D IOXLG ZKHQHYHU WKHUH LV D VKHDULQJ IORZ DQG WKH LQHUWLDO HIIHFWV DUH PXFK ODUJHU WKDQ WKH YLVFRXV HIIHFWV /XPOH\ DQG 3DQRIVN\ f RSWHG WR OLVW WKH HVVHQWLDO IHDWXUHV RI WXUEXOHQFH 7KHLU GHVFULSWLRQ FDQ EH VXPPHG XS DV 7XUEXOHQFH LV D PRGH RI IOXLG PRWLRQ WKDW LV D WKUHHGLPHQVLRQDO FRQWLQXXP SKHQRPHQRQ URWDWLRQDO GLVVLn SDWLYH QRQOLQHDU VWRFKDVWLF DQG GLIIXVLYH ZLWK WKH WUDQVSRUW RFFXUULQJ DW WLPH DQG OHQJWK VFDOHV W\SLFDOO\ WKH VDPH DV WKRVH RI WKH SURSHUWLHV EHLQJ WUDQVIHUUHG 2WKHU LQWURGXFWRU\ GLVFXVVLRQV RI WXUEXOHQFH DUH JLYHQ E\ 6XWWRQ f 6FKOLFKWLQJ f 7RZQVHQG f %URGNH\ f %DWFKHORU f 3KLOOLSV f 0RQLQ DQG
PAGE 19
WKH FRUUHODWLRQ EHWZHHQ WKH YHORFLWLHV DW WZR SRLQWV DV RQH RI WKH TXDQWLWLHV QHHGHG WR GHVFULEH WXUEXOHQFH +LV WKHRU\ DOVR FOHDUO\ VKRZV WKH HQRUPRXV VLPSOLILFDWLRQV REWDLQHG E\ DVVXPLQJ KRPRJHQHLW\ DQG LVRWURS\ 7KLV DVVXPSWLRQ KDV EHHQ ZLGHO\ XVHG VLQFH WR DFKLHYH VLPSOLILFDWLRQV LQ IRUPXODWLRQ RI WKH WXUEXOHQFH WKHRULHV 5LFKDUGVRQ f VXJJHVWHG WKDW WKH HQHUJ\ IURP WKH PHDQ IORZ LV GUDZQ E\ WKH ODUJH HGGLHV DQG WKH ODUJH HGGLHV LQ WXUQ EUHDN XS LQWR VPDOOHU HGGLHV XQWLO WKH VPDOOHVW VL]H RI HGGLHV LH WKH GLVVLSDWLRQ VFDOH LV UHDFKHG .HOOHU DQG )ULHGPDQQ LQ ZHUH WKH ILUVW WR VWDWH WKDW WKH FRUUHODWLRQ IXQFWLRQV DQG WKH VWDWLVWLFDO PRPHQWV DUH WKH IXQGDPHQWDO FKDUDFWHULVWLFV RI WXUEXOHQFH 0RQLQ DQG
PAGE 20
7D\ORU Df VKRZHG WKH H[LVWHQFH RI LQWHUDFWLRQV EHWZHHQ FRPSRQHQWV KDYLQJ GLIIHUHQW OHQJWK VFDOHV 7D\ORU Ef VWXGLHG WKH HQHUJ\ VSHFWUXP IXQFWLRQ RI WXUEXOHQFH 7KH HQHUJ\ VSHFWUXP IXQFWLRQ GHVFULEHV WKH GLVWULEXWLRQ RI WXUEXOHQW NLQHWLF HQHUJ\ RYHU YDULRXV HGG\ VL]HV +H VKRZHG WKDW WKLV HQHUJ\ VSHFWUXP IXQFWLRQ LV WKH )RXULHU WUDQVIRUP RI WKH FRUUHODWLRQ EHWZHHQ WKH HGG\ YHORFLWLHV DW WZR SRLQWV 9RQ .DUP£Q D DQG Ef VKRZHG WKDW WKH YHORFLW\ FRUUHODWLRQV DUH WHQVRUV 7KLV KDV HQDEOHG PRUH SUHFLVH DQG FRQFLVH ZRUNLQJ HTXDWLRQV IRU WKH QRQLVRWURSLF VLWXDWLRQV 9RQ .DUP£Q DQG +RZDUWK f UHGXFHG WKHVH WHQVRUV WR VFDODU TXDQWLWLHV LQ WKH FDVH RI LVRWURSLF WXUEXOHQFH 7KLV H[HPSOLILHV WKH VLPSOLILFDWLRQ DFKLHYHG WKURXJK WKH DVVXPSWLRQ RI LVRWURS\ .ROPRJRURY LQ SXW IRUZDUG WKH WKHRU\ RI XQLYHUVDO HTXLOLEULXP ,W VWDWHV WKDW WKH WXUEXOHQFH HQHUJ\ LV JHQHUDWHG LQ D FHUWDLQ UDQJH RI OHQJWK VFDOH DQG WKHQ SDVVHG RQ WR VPDOOHU DQG VPDOOHU VFDOHV E\ D FDVFDGH SURFHVV 7KXV WKH VL]HV DUH RQO\ LQn GLUHFWO\ FRXSOHG $W KLJK 5H\QROGV QXPEHUV WKH FUHDWLRQ DQG WKH GLVVLSDWLRQ UDQJHV DUH ZLGHO\ VHSDUDWHG (OOLVRQ f 2EXNKRY f DUULYHG DW WKH VDPH FRQFOXVLRQV LQGHSHQGHQWO\ .ROPRJRURY DUULYHV DW WKH FRQFHSW RI ORFDO LVRWURS\ $V WKH VL]HV RI WKH HGGLHV JHW VPDOOHU WKHUH LV D WHQGHQF\ IRU WKH WXUEXOHQFH VWUXFWXUH WR EH LVRWURSLF .ROPRJRURYf¬V WKHRU\ OHDGV WR WKH FRQFOXVLRQ 7HQQHNHV DQG /XPOH\ f WKDW WKH WXUEXOHQFH HQHUJ\ LQ WKH RYHUODS UHJLRQ RI WKH HQHUJ\ FUHDWLRQ UDQJH DQG WKH .ROPRJRURY UDQJH FDQ EH H[SUHVVHG DV
PAGE 21
(Nf D H N f ZKHUH ( LV WKH HQHUJ\ VSHFWUXP IXQFWLRQ e LV WKH UDWH RI WXUEXOHQFH HQHUJ\ GLVVLSDWLRQ DQG N LV WKH ZDYH QXPEHU 7KLV UDQJH LV NQRZQ DV WKH LQHUWLDO VXEUDQJH .ROPRJRURY f DQG 2EXNKRY f UHILQHG WKHLU WKHRU\ RI WKH VPDOO VFDOH VWUXFWXUH GHSHQGLQJ XSRQ RQH YDOXH RI WKH HQHUJ\ GLVVLSDWLRQ e DQG Yf RQO\ ,Q IDFW e LV D IOXFWXDWLQJ TXDQWLW\ DQG WKH PHDQ e YDULHV ZLWK WKH 5H\QROGV QXPEHU $ VOLJKW GHYLDWLRQ IURP (T f LV REVHUYHG E\ WDNLQJ LQWR DFFRXQW WKH YDULDWLRQV LQ WKH HQHUJ\ GLVVLSDWLRQ .UDLFKQDQ f SURSRVHG D VROXWLRQ WR WKH FORVXUH SUREOHP E\ H[SDQGLQJ WKH WKLUG DQG KLJKHU RUGHU VWDWLVWLFDO PRPHQWV RI WKH )RXULHU WUDQVIRUP RI WKH YHORFLW\ LQ WHUPV RI WKH VHFRQG RUGHU PRPHQWV DQG WKHLU UHVSRQVH IXQFWLRQV 7KH UHVSRQVH IXQFWLRQ RI WKH WLPH FRUUHODWLRQ FRHIILFLHQW ZLWK WKH ODJ WLPH [ LV H[S 9N [f ZKHUH N LV WKH ZDYH QXPEHU +H DSSOLHG WKH WKHRU\ WR VWDWLRQDU\ LVRWURSLF WXUEXOHQFH RI YHU\ KLJK 5H\QROGV QXPEHU ZKHUH WKH 5H\QROGV QXPEHU ZDV GHILQHG LQ WHUPV RI WKH ZDYH QXPEHU DQG WKH WXUEXOHQW LQWHQVLW\ 7KH HQHUJ\ VSHFWUXP IXQFWLRQ LQ WKLV UDQJH ZDV REWDLQHG LQ WHUPV RI D QRQOLQHDU LQWHJUDO LQYROYLQJ WKH WLPH FRUUHODWLRQ IXQFWLRQ DQG UHVSRQVH IXQFWLRQV 9DULRXV FORVXUH WKHRULHV KDYH EHHQ SXW IRUZDUG E\ +HLVHQEHUJ f .RYDV]QD\ f 3URXGPDQ DQG 5HLG f &KDQGUDVHNKDU f 7DWVXPL f 3DWDQNDU DQG 6SDOGLQJ f DQG RWKHUV $ VXUYH\ RI GHYHORSPHQWV LQ WKH ILHOG XS WR LV GLVFXVVHG E\ 5H\QROGV f
PAGE 22
7XUEXOHQFH PHDVXUHPHQWV 2QH RI WKH PRUH VLJQLILFDQW VHWV RI H[SHULPHQWDO PHDVXUHPHQWV ZDV PDGH E\ /DXIHU f IRU DLU LQ IXOO\ GHYHORSHG IORZ LQ D SLSH DW 5H\QROGV QXPEHUV IURP DERXW WR 7KH PHDQ IORZ DQG WKH VWDWLVWLFDO SURSHUWLHV PHDVXUHPHQWV LQFOXGH WKH PHDQ YHORFLW\ SURILOH UDGLDO D[LDO DQG DQJXODU WXUEXOHQFH LQWHQVLWLHV FURVVn FRUUHODWLRQV EHWZHHQ WKH GLIIHUHQW FRPSRQHQWV RI WKH IOXFWXDWLQJ YHORFLW\ HQHUJ\ VSHFWUD DQG WKH WXUEXOHQW NLQHWLF HQHUJ\ GLVVLSDn WLRQ UDWH ,Q WKH DWPRVSKHUH 5LFKDUGVRQ f GLG QRW XVH DQ\ RI WKH UHILQHG LQVWUXPHQWV DYDLODEOH WRGD\ EXW ZDV DEOH WR PDNH VRPH PHDQLQJIXO DSSUR[LPDWLRQV DQG DUULYH DW XVHIXO UHVXOWV 7D\ORU f VWXGLHG WKH EHKDYLRU RI LVRWURSLF WXUEXOHQFH LQ D GHFD\LQJ ZLQG VWUHDP 7D\ORU Ef XVLQJ D KRW ZLUH DQHPRPHWHU PHDVXUHG WKH NLQHWLF HQHUJ\ DVVRFLDWHG ZLWK GLIIHUHQW ZDYH QXPEHUV LH WKH HQHUJ\ VSHFWUD 'HLVVOHU f VWXGLHG DGLDEDWLF WXUEXOHQW IORZV LQ SLSHV )DYUH HW DO f PHDVXUHG VSDFH DQG WLPH FRUUHODWLRQ IXQFWLRQV LQ D ZLQG WXQQHO DQG D ERXQGDU\ OD\HU /DXIIHU f FRQGXFWHG D GHWDLOHG VWXG\ RI WKH PHDQ DQG IOXFWXDWLQJ TXDQWLWLHV LQ D WZR GLPHQVLRQDO FKDQQHO +H PDGH WKH HQHUJ\ EDODQFHV DQG YHULILHG .ROPRJRURYnV K\SRWKHVLV 7KH LVRWURSLF WXUEXOHQFH ZDV WKH REMHFW RI WKH VWXGLHV RI /LHSPDQQ DQG /DXIIHU f +DQUDWW\ f DQG &RUUVLQ f ,Q D VWXG\ RI WXUEXOHQFH LQ D ERXQGDU\ OD\HU RYHU D IODW SODWH .OHEDQRII DQG 'LHKO f DQG .OHEDQRII f IRXQG WKDW WKH
PAGE 23
SUREDELOLW\ GHQVLWLHV RI WKH IOXFWXDWLQJ YHORFLWLHV DUH YHU\ FORVH WR *DXVVLDQ 2WKHU LPSRUWDQW FRQWULEXWLRQV LQ WKH PHDVXUHPHQWV RI ODERUDWRU\ VFDOH WXUEXOHQFH LQFOXGH WKH ZRUNV RI 5RWWD f /LHSPDQQ DQG 5RELQVRQ f 3DL f 6DQGERUQ f 6DQGERUQ DQG %UDXQ f +DQUDWW\ f )UHQNLHO f -RKQN DQG +DQUDWW\ DEf %DUELQ DQG -RQHV f &HUPDN DQG %DOGZLQ f %DOGZLQ DQG 0LFNHOVHQ f &ODUN f *UDVV f DQG 6Q\GHU DQG /XPOH\ f 6KHSSDUG f PHDVXUHG WKH YHORFLW\ SURILOH LQ WKH DWPRVSKHUH 6KHSSDUG XVHG D GUDJ SODWH DW WKH JURXQG WR PHDVXUH WKH GUDJ IRUFH 7KH YHORFLW\ SURILOH ZDV REVHUYHG WR EH ORJDULWKPLF IRU QHDU QHXWUDO FRQGLWLRQV X I ]] ! f ZKHUH . LV WKH YRQ .DUPDQ FRQVWDQW 7 LV WKH VKHDU VWUHVV RQ WKH R JURXQG S LV WKH GHQVLW\ RI DLU DQG ]T LV WKH URXJKQHVV OHQJWK 7KH YDOXH RI WKH GUDJ FRHIILFLHQW DV D IXQFWLRQ RI WKH URXJKQHVV OHQJWK ZDV HYDOXDWHG DQG WKH YDULDWLRQ RI WKH PL[LQJ OHQJWK ZLWK KHLJKW ZDV VWXGLHG )RU RQH REVHUYDWLRQ VLWH 7DLW f REWDLQHG DQ HPSLULFDO UHODWLRQVKLS IRU WKH WHPSHUDWXUH JUDGLHQW LQ WHUPV RI WKH LQWHQVLW\ RI UDGLDWLRQ /HWWDX f H[DPLQHG WKH HIIHFW RI VXUIDFH URXJKQHVV RQ WKH JHRVWURSKLF GUDJ FRHIILFLHQW 7KH JHRVWURSKLF GUDJ FRHIILFLHQW LV WKH UDWLR RI ZLQG VSHHG WR WKH JHRVWURSKLF VSHHGf§ZKHUH WKH FRULROLV IRUFH EDODQFHV WKH SUHVVXUH JUDGLHQW IRUFH +H DOVR GLVFXVVHG WKH
PAGE 24
HIIHFW RI )LFKDUGVRQ QXPEHU RQ WKH VKHDU VWUHVV DQG WKH YHORFLW\ SURILOH 3DQRIVN\ DQG 'HODQG f DQDO\]HG WKH RQHGLPHQVLRQDO HQHUJ\ VSHFWUD LQ WKH ORZHVW RQH KXQGUHG PHWHUV RI WKH DWPRVSKHUH 7KH\ IRXQG WKH 7D\ORU K\SRWKHVLV UHJDUGLQJ HTXLYDOHQFH RI (XOHULDQ WLPH DQG VSDFH FRUUHODWLRQV WR EH FRUUHFW DQG REVHUYHG WKDW WKH /DJUDQJLDQ DQG (XOHULDQ FRUUHODWLRQV KDYH VLPLODU VKDSHV 5HFRUG DQG &UDPHU f FRPSDUHG WKH HQHUJ\ GLVVLSDWLRQ UDWHV REWDLQHG E\ SUHYLRXV LQYHVWLJDWRUV DQG IRXQG WKDW WKH YDULDWLRQV LQ H DQG ZLWK VWDELOLW\ DQG KHLJKW ZHUH H[WUHPHO\ ODUJH 7KH WXUEXOHQFH LQWHQVLW\ DQG WKH HQHUJ\ VSHFWUD GDWD PHDVXUHG DW 5RXQG +LOO ZHUH FRUUHODWHG LQ WKH ODWHU VWXGLHV &UDPHU HBW DOB &UDPHU DQG 5HFRUG f 2WKHU PHDVXUHPHQWV RI WKH WXUEXOHQW LQWHQVLWLHV DQG WKH HQHUJ\ VSHFWUD LQFOXGH WKH ZRUNV RI 7D\ORU f *LIIRUG f 3DQRIVN\ MLW DO f 7KRPSVRQ f DQG $QJHO¯ f 7KH VWXG\ RI IOX[JUDGLHQW UHODWLRQV DQG WKH WHPSHUDWXUH IOXFWXDWLRQV LQ WKH JURXQG OD\HU KDYH EHHQ FRQGXFWHG E\ /HWWDX f &RUUVLQ f %ROJLDQR f 3ULHVWOH\ f /XPOH\ f :HEVWHU f '\HU f *XUYLFK DQG =XENRYVNLL f &ODUNH f 2NDPDWR DQG :HEE f DQG +DXJHQ HW DO f +DXJHQ HW DO f H[SHULPHQWDOO\ IRXQG WKDW PRPHQWXP DQG KHDW IOX[HV DUH FRQVWDQW WR ZLWKLQ b SURYLGHG WKDW D VXIILFLHQWO\ ORQJ DYHUDJLQJ SHULRG LV FKRVHQ 7KH YHORFLW\YHORFLW\ DQG YHORFLW\ WHPSHUDWXUH FRUUHODWLRQV ZHUH PHDVXUHG DQG LW ZDV VKRZQ WKDW WKH YHUWLFDO YHORFLW\ IOXFWXDWLRQ Zn LV EHWWHU FRUUHODWHG ZLWK WHPSHUDWXUH WKDQ ZLWK WKH KRUL]RQWDO YHORFLW\ IOXFWXDWLRQ Xn
PAGE 25
7KH VLPXODWLRQ RI WKH DWPRVSKHULF JURXQG OD\HU LQ D ZLQG WXQQHO KDV EHHQ RI VRPH LQWHUHVW WR WKH H[SHULPHQWDOLVWV LQ UHFHQW WLPHV 6WURP DQG +DOLWVN\ f RXWOLQH YDULRXV IDFWRUV WR EH FRQVLGHUHG LQ GHVLJQLQJ D ZLQG WXQQHO /OR\G f JHQHUDWHG D VKHDULQJ IORZ LQ D ZLQG WXQQHO ZLWK DQ DUWLILFLDOO\ URXJKHQHG IORRU 3UXSSDFKHU DQG 1HLOEXUJHU f GLVFXVV WKH GHVLJQ RI D ZLQG WXQQHO DW WKH 8QLYHUVLW\ RI &DOLIRUQLD $U\D f DQG 6FKRQ HWB DO f VLPLODWHG WKHUPDOO\ VWUDWLILHG ERXQGDU\ OD\HUV LQ ZLQG WXQQHOV DQG FRPSDUHG WKHLU REVHUYDn WLRQV ZLWK WKH DWPRVSKHULF GDWD &HUPDN HWB DO f SHUIRUPHG D GLPHQVLRQDO DQDO\VLV IRU D ZLQG WXQQHO DQG QXPHURXV H[SHULPHQWV 7KH 0RQLQ2EXNKRY VLPLODULW\ WKHRU\ 7KH VLPLODULW\ WKHRU\ RI 0RQLQ DQG 2EXNKRY f E\ GLPHQVLRQDO DQDO\VLV DUULYHV DW FKDUDFWHULVWLF OHQJWK DQG WHPSHUDWXUHV LQ D WKHUn PDOO\ VWUDWLILHG JRXQG OD\HU 9HUWLFDO WHPSHUDWXUH JUDGLHQWV LQ WKH DWPRVSKHUH FUHDWH GHQVLW\ GLIIHUHQFHV ODUJH HQRXJK IRU WKH EXR\DQF\ WHUP WR EH LQFOXGHG LQ WKH HTXDWLRQ RI PRWLRQ +RZHYHU EDVHG RQ HPSLULFDO REVHUYDWLRQV WKH YHUWLFDO KHDW DQG PRPHQWXP IOX[HV FDQ EH DVVXPHG WR EH FRQVWDQW LQ WKH JURXQG OD\HU 7KH RWKHU DVVXPSWLRQV WKDW PXVW EH PDGH WR GHULYH WKH VLPLODULW\ WKHRU\ DUH Lf FRPSOHWH KRPRJHQHLW\ H[LVWV LQ WKH KRUL]RQWDO [\f SODQH LLf YDULDWLRQV RI GHQVLW\ SURGXFHG E\ SUHVVXUH IOXFWXDWLRQV DUH QHJOHFWHG LLLf WKH WHPSHUDWXUH GHYLDWLRQV IURP WKH QHXWUDO FRQGLWLRQV DUH VPDOO DQG LYf WKH YLVFRXV HIIHFWV DUH QHJOHFWHG 7KHQ WKH WKUHH FRPSRQHQWV RI WKH HTXDWLRQ RI PRWLRQ WKH HTXDWLRQ RI FRQWLQXLW\ DQG WKH WKHUPDO HQHUJ\ HTXDWLRQ UHGXFH WR
PAGE 26
'X B 3 'W S [ 'Y B 3 'W S \ f f 'Z " 7B 'W S ] 7G . f X Y BZ [ \ ] f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r DQG T&AS ZKHUH Xr LV WKH IULFWLRQ YHORFLW\ 7KH RQO\ OHQJWK DQG WHPSHUDWXUH VFDOHV WKDW FDQ EH IRUPHG E\ D FRPELQDWLRQ RI WKHVH SDUDPHWHUV DUH / X e BB 75 &S3 f 7 r f
PAGE 27
,I WKH VLPLODULW\ WKHRU\ KROGV WKH GHSHQGHQFH RI DQ\ PHDQ WXUEXOHQFH FKDUDFWHULVWLF ) RQ KHLJKW PD\ EH ZULWWHQ LQ WKH IRUP 9 ]/f f R ZKHUH )T LV D SDUDPHWHU ZLWK WKH GLPHQVLRQV RI ) IRUPHG E\ FRPELQDWLRQ RI J74 3! Xr DQG T&AS DQG LS LV D XQLYHUVDO IXQFWLRQ RI ]/ 3DQRIVN\nV f DQDO\VLV EDVHG RQ WKH PHDVXUHPHQWV RI 3ULHVWOH\ f VKRZV WKDW WKH VLPLODULW\ WKHRU\ KROGV IRU WKH YHORFLW\ DQG WHPSHUDWXUH SURILOHV 7KH VLPLODULW\ WKHRU\ ZDV XVHG E\ &KDUQRFN f WR FRUUHODWH 5LFKDUGVRQ QXPEHU GLPHQVLRQOHVV YHORFLW\ DQG WHPSHUDWXUH JUDGLHQWV DV IXQFWLRQV RI WKH GLPHQVLRQOHVV KHLJKW ]/ 7KH 5LFKDUGVRQ QXPEHU FDQ EH LQWHUSUHWHG DV WKH JUDYLWDn WLRQDO IRUFH RYHU WKH LQHUWLDO IRUFH 7D\ORU f .OXJ f DQG 6ZLQEDQN f FRPSDUHG WKH REVHUYDWLRQV PDGH LQ WKH DWPRVSKHUH ZLWK WKH VLPLODULW\ WKHRU\ &DOGHU f LV FULWLFDO RI WKH VLPLODULW\ WKHRU\ DQG VD\V WKDW WKH VLPLODULW\ WKHRU\ LV QRW D VWULFW FRQVHTXHQFH RI G\QDPLFDO HTXDWLRQV 7KHUH LV QR FORVHG V\VWHP RI HTXDWLRQV IRU DQ\ ILQLWH QXPEHU RI VWDWLVWLFDO FKDUDFWHULVWLFV RI WXUEXOHQFHf§QRW HYHQ IRU WKH YHORFLW\ DQG WKH WHPSHUDWXUH JUDGLHQWV 7KH K\SRWKHVLV RI WKH H[LVWHQFH RI RQO\ WKUHH SDUDPHWHUV J7 Xr T& S LV D VSHFLILF YLHZSRLQW RQ . [ S WKH PHFKDQLVP RI WXUEXOHQFH DQG LV D K\SRWKHVLV RI PD[LPXP VLPSOLFLW\ $WPRVSKHULF 'LIIXVLRQ (DUO\ GHYHORSPHQWV DQG WKH .WKHRU\ 7KH WLPH DYHUDJHG HTXDWLRQ RI FRQWLQXLW\ IRU D VSHFLHV LQ D
PAGE 28
PHGLXP FDQ EH H[SUHVVHG LQ JHQHUDO DV %LUG HWB DO f QU YYnFn '9 F U f ZKHUH F LV WKH WLPH DYHUDJHG FRQFHQWUDWLRQ RI WKH VSHFLHV 9Bn LV WKH IOXFWXDWLQJ YHORFLW\ YHFWRU ' LV WKH PROHFXODU GLIIXVLYLW\ DQG U LV WKH UDWH RI SURGXFWLRQ RI WKH VSHFLHV E\ FKHPLFDO UHDFWLRQ $ VLPSOH IRUP RI (T f LV REWDLQHG E\ PDNLQJ WKH IROORZLQJ DVVXPSWLRQV LfWKH WXUEXOHQW PDVV IOX[ FDQ EH H[SUHVVHG LQ D PDQQHU DQDORJRXV WR WKDW XVHG E\ %RXVVLQHVT f WR GHVFULEH WKH WXUEXOHQW PRPHQWXP IOX[ LH LW LV WDNHQ DV EHLQJ SURSRUWLRQDO WR WKH SRWHQWLDO JUDGLHQW RU 9nFr .r9F f LLfPROHFXODU GLIIXVLRQ LV QHJOHFWHG LLLfFKHPLFDO UHDFWLRQ LV QHJOHFWHG 5REHUWV f XVHG (T f DVVXPLQJ WKH HGG\ GLIIXVLYLW\ WR EH D VFDODU TXDQWLW\ ZKLFK LPSOLHV LVRWURS\f WKDW LV LQGHSHQGHQW RI SRVLWLRQ IRU WKH FDVHV RI GLVSHUVLRQ RI D SROOXWDQW IURP Lf DQ LQVWDQWDQHRXV SRLQW VRXUFH LLf D FRQWLQXRXV SRLQW VRXUFH LQ D XQLIRUP ZLQG DQG LLLf D FRQWLQXRXV LQILQLWH OLQH VRXUFH SHUSHQGLFXODU WR WKH GLUHFWLRQ RI D XQLIRUP ZLQG +H DOVR REWDLQHG DQ H[SUHVVLRQ IRU WKH YLVLEOH RXWOLQH RI D VPRNH SOXPH DQG H[WHQGHG WKH UHVXOWV REWDLQHG ZLWK DQ LVRWURSLF GLIIXVLYLW\ WR DQLVRWURS\
PAGE 29
5LFKDUGVRQ f FRQVLGHUHG WKH GLIIXVLRQ SURFHVV DV WKH VFDWWHU RI D FOXVWHU RI SROOXWDQW SDUWLFOHV +H XVHG D VFDODU GLI IXVLYLW\ WKDW ZDV LQGHSHQGHQW RI WKH SRVLWLRQ LQ VSDFH EXW ZDV D IXQFWLRQ RI WKH HIIHFWLYH HGG\ VL]H E\ FRQVLGHULQJ WKH GLVWDQFH EHWZHHQ WKH GLIIXVLQJ DQG WKH QHLJKERULQJ SDUWLFOHV 6XWWRQ f GHYHORSHG WKH LGHD RI WKH HIIHFWLYH HGG\ VL]H +H VWDUWHG ZLWK 7D\ORUnV f WKHRU\ D Wr 5OWf G7 GWn f ZKHUH D LV WKH YDULDQFH RI WKH GLVSHUVLRQ 2 LV WKH YDULDQFH RI WKH Y YHORFLW\ DQG 5A LV WKH /DJUDQJLDQ FRUUHODWLRQ FRHIILFLHQW +H DVVXPHG WKDW 97f 9 Y D W 9 Q f Q ! DQG GHULYHG DQ H[SUHVVLRQ IRU D +LV UHVXOW VKRZHG WKDW IRU Q WKH VL]H RI WKH HIIHFWLYH HGG\ LQFUHDVHV ZLWK WKH GLVWDQFH RI WKH GLIIXVLQJ FOXVWHU IURP LWV VRXUFH 6XWWRQ f GHYHORSHG DQ H[SUHVVLRQ IRU WKH PDVV IOX[ GXH WR HYDSRUDWLRQ IURP D VXUIDFH DVVXPLQJ WKDW WKH GLIIXVLRQ RI PDVV IROORZV WKH VDPH ODZV DV WKH GLIIXVLRQ RI PRPHQWXP 6XWWRQ f H[WHQGHG KLV SUHYLRXV WKHRU\ RI HGG\ GLIIXVLRQ WR LQFOXGH WKH WXUEXOHQFH LQWHQVLWLHV DQG H[WHQGHG WKH DQDO\VLV WR KHDW WUDQVIHU DV ZHOO 6XWWRQ f
PAGE 30
%RVDQTXHW DQG 3HDUVRQ f SURSRVHG D VROXWLRQ RI (T f ZKLFK LV ZHOO NQRZQ DV WKH .WKHRU\ 9HORFLW\ LV DVVXPHG LQGHSHQGHQW RI KHLJKW DQG WKH HGG\ GLIIXVLYLWLHV LQ WKH \ DQG WKH ]GLUHFWLRQV DUH DVVXPHG DV IROORZV . T X [ \ f . T X ] ] f ZKHUH DQG DUH WKH HGG\ PDVV GLIIXVLYLW\ FRPSRQHQWV LQ WKH \ DQG ]GLUHFWLRQV UHVSHFWLYHO\ TA DQG Te DUH FRQVWDQWV X LV WKH FRQVWDQW PHDQ ZLQG YHORFLW\ DQG [ DQG ] DUH WKH PHDQ ZLQG DQG WKH YHUWLFDO GLVWDQFHV UHVSHFWLYHO\ &DOGHU f DSSUR[LPDWHG WKH ORJDULWKPLF YHORFLW\ SURILOH IRU QHXWUDO FRQGLWLRQV E\ WKH SRZHU ODZ SURILOH DQG REWDLQHG WKH VROXn WLRQ RI WKH WZRGLPHQVLRQDO GLIIXVLRQ HTXDWLRQ &DOGHU f PDGH D FRQWULEXWLRQ WR WKH .WKHRU\ E\ H[SUHVVLQJ WKH HGG\ GLIIXVLYLW\ DV . . X ] f ZKHUH Xr LV WKH IULFWLRQ YHORFLW\ DQG . LV WKH YRQ .DUPDQ FRQVWDQW 6WDUWLQJ ZLWK D SRZHU ODZ IRUP RI WKH YHORFLW\ SURILOH DQG DVVXPLQJ WKH WXUEXOHQW 6FKPLGW QXPEHU WR EH RQH 'HDFRQ f DUULYHV DW WKH IROORZLQJ IRUP IRU WKH YHUWLFDO WXUEXOHQW GLIIXVLYLW\ . . 8M ] D f D ! XQVWDEOH FRQGLWLRQV G VWDEOH FRQGLWLRQV
PAGE 31
6PLWK f DOVR XVHG D SRZHU ODZ IRUP IRU WKH YHUWLFDO HGG\ GLIIXVLYLW\ DQG DVVXPHG WKDW WKH HGG\ GLIIXVLYLW\ LQ WKH \ GLUHFWLRQ DOVR YDULHV LQ DQ LGHQWLFDO PDQQHU ZLWK KHLJKW +H REWDLQHG H[SUHVVLRQV IRU JURXQG OHYHO FRQFHQWUDWLRQ SURILOHV IURP HOHYDWHG VRXUFHV )XUWKHU GHYHORSPHQWV RI WKH .WKHRU\ LQFOXGH WKH ZRUNV RI 'DYLHV f .QLJKWLQJ f DQG 5RXQGV f *DXVVLDQ SOXPH PRGHOV +D\ DQG 3DVTXLOO f PHDVXUHG GLVSHUVLRQ GRZQ ZLQG RI D VRXUFH DQG VKRZHG WKDW WKH /DJUDQJLDQ FRUUHODWLRQV DUH PDLQWDLQHG KLJK IRU PXFK ORQJHU SHULRGV RI WLPH WKDQ WKH (XOHULDQ FRUUHODWLRQV 7KLV ZRUN IRUPHG WKH EDVLV DQG ODWHU OHG WR WKH GHYHORSPHQW RI WKH *DXVVLDQ SOXPH PRGHOV +D\ DQG 3DVTXLOO f H[SORUHG WKH LGHD IXUWKHU DQG K\SRWKHVL]HG WKDW WKH /DJUDQJLDQ DQG WKH (XOHULDQ FRUUHODn WLRQV DUH VLPLODU LQ VKDSHV EXW KDYH GLIIHUHQW WLPH VFDOHV ,PSO\LQJ D FRUUHODWLRQ FRHIILFLHQW HTXDO WR RQH WKH\ DVVXPHG &-\ D ; f ZKHUH D LV WKH VWDQGDUG GHYLDWLRQ RI WKH FRQFHQWUDWLRQ GLVWULEXWLRQ \ LQ WKH \GLUHFWLRQ FURVV ZLQG GLUHFWLRQf 2J LV WKH VWDQGDUG GHYLDWLRQ RI WKH KRUL]RQWDO ZLQG GLUHFWLRQ DQG [ LV WKH GRZQ ZLQG GLVWDQFH +RZHYHU WKH FRUUHODWLRQ FRHIILFLHQW LQ SUDFWLFH LV DOZD\V OHVV WKDQ RQH DQG (T f VKRXOG EH PRGLILHG DFFRUGLQJO\ %DVHG RQ WKH DYDLODEOH GDWD +DXJHQ 0RQLQ 6PLWK DQG +D\ DQG D IHZ RWKHUVf 3DVTXLOO f SORWWHG YV [ IRU GLIIHUHQW &7T YDOXHV 6LPLODU FXUYHV IRU YV [ ZHUH DOVR SORWWHG ,Q WKH VDPH
PAGE 32
SDSHU 3DVTXLOO DOVR VKRZHG KRZ WR HVWLPDWH J IURP TXDOLWDWLYH PHWHRURORJLFDO SDUDPHWHUV *LIIRUG f VKRZHG KRZ VLPLODULW\ SUHGLFWLRQV RI UHODWLYH GLIIXVLRQ WKHRU\ %DWFKHORU f DSSO\ WR WKH VSUHDGLQJ RI VPRNH SXIIV DQG XVLQJ WKH REVHUYDWLRQV RI .HOORJ f DQG )UHQNLHO DQG .DW] f FRUUHODWHG PHDQ VTXDUH SDUWLFOH GLVSHUVLRQ DV D IXQFn WLRQ RI WLPH *LIIRUG f HVWLPDWHG WKH GLIIXVLRQ FRHIILFLHQWV LQ 6XWWRQnV f IRUPXODWLRQ DQG RXWOLQHG WKH PHWKRG WR HVWLPDWH WKH SDUDPHWHUV QHHGHG WR XVH 3DVTXLOOnV f JUDSKV 3DVTXLOO DQG *LIIRUG ERWK ZRUNHG LQGHSHQGHQWO\ WR HVWLPDWH WKH nV DQG WKH FKDUWV RULJLQDOO\ IRUPXODWHG E\ 3DVTXLOO DUH EHWWHU NQRZQ DV 3DVTXLOO*LIIRUG FKDUWV %RZQH f KDV SUHVXPDEO\ LPSURYHG WKH FKDUWV E\ LQFOXGLQJ WKH HIIHFW RI WKH GRZQ ZLQG WHUUDLQ LQ D TXDOLWDWLYH PDQQHU 7KH *DXVVLDQ SOXPHV DUH ZLGHO\ XVHG LQ LQGXVWU\ DQG E\ WKH SROOXWLRQ FRQWURO DJHQFLHV )UHQNLHO D 'DYLGVRQ &URNH 6KLHK f 6RPH UHFHQW VWXGLHV 7KH QXPEHU DQG W\SHV RI WKH YDULDEOHV LQYROYHG LQ DQ DWPRVSKHULF GLVSHUVLRQ SUREOHP PDNH WKH SUREOHP D YHU\ FRPSOH[ RQH LQ QDWXUH 3DQRIVN\ f VXPPDUL]HV WKH GLIIHUHQW PHWHRURORJLFDO DQG SK\VLFDO SDUDPHWHUV LQYROYHG DQG WKHLU HIIHFWV 6HLQIHOG f LQFOXGHV WKH HIIHFW RI FKHPLFDO UHDFWLRQV DQG VXUYH\V WKH GLIIHUHQW GLIIXVLRQ PRGHOV ,QWURGXFWRU\ ERRNV RQ WKH VXEMHFW DUH WKRVH RI 6XWWRQ f 3ULHVWOH\ f 3DVTXLOO f 6ODGH f &VDQDG\ f DQG 6HLQIHOG f 7KH HPSKDVLV LQ WKH FXUUHQW UHVHDUFK ZRUN KDV
PAGE 33
VKLIWHG WR WKH SUHGLFWLRQ RI GLVSHUVLRQ UDWKHU WKDQ SUHVHQWDWLRQ RI WKHRULHV 7KLV KDV OHG WR D FHUWDLQ DPRXQW RI ILHOG PHDVXUHPHQWV DQG WR WKH HVWLPDWLRQ RI HGG\ GLIIXVLYLWLHV 6WHZDUW H8 DO f XVHG UDGLRDFWLYH DUJRQ DV D WUDFHU PDWHULDO IURP D P VWDFN %(32 UHDFWRUf WR VWXG\ WKH GLIIXVLRQ RI WKH SOXPH LQ WKH DWPRVSKHUH GRZQ ZLQG RI WKH VWDFN 7KH HIIHFWLYH YDOXHV RI WKH FRQVWDQWV LQ 6XWWRQnV f GLIIXVLYLW\ H[SUHVVLRQV ZHUH HYDOXDWHG 6DIIPDQ f VWXGLHG WKH HIIHFW RI ZLQG VKHDU RQ KRUL]RQWDO VSUHDG IURP DQ LQVWDQWDQHRXV JURXQG VRXUFH ,VOLW]HU DQG 'XPEDXOG f VKRZHG WKDW WKH GHSRVLWLRQ RI SDUWLFXODWH SROOXWDQWV FDQ EH VLJQLILFDQW WKXV UHGXFLQJ FRQFHQWUDWLRQV DW JUHDWHU GLVWDQFHV WR ZHOO EHORZ WKDW SUHGLFWHG E\ GLIIXVLRQ PRGHOV )XTXD\ HW DO f *UHHQ *ORZ SURJUDPf FDUULHG RXW JURXQG VRXUFH H[SHULPHQWV DQG FRQFOXGHG WKDW WKH VWDQGDUG GHYLDWLRQ RI WKH ZLQG GLUHFWLRQ *T LV QRW D SDUDPHWHU WR XVH IRU FURVV ZLQG GLIIXVLRQ EXW WKH SURGXFW 2T8 VKRXOG EH XVHG LQVWHDG %ODFNDGDU f UHODWHG WKH WXUEXOHQW PRPHQWXP GLIIXVLYLW\ . WR ZLQG VKHDU WKURXJK WKH XVH RI WKH H[SUHVVLRQ . e e f P ZKHUH e LV D OHQJWK VFDOH RI WXUEXOHQFH 6WDUWLQJ IURP WKH WXUEXOHQW HQHUJ\ HTXDWLRQ WKH HQHUJ\ GLVVLSDWLRQ e ZDV H[SUHVVHG LQ WHUPV RI WKH ZLQG VKHDU IRU QHXWUDO FRQGLWLRQV :X f H[WHQGHG WKH ZRUN IRU QRQQHXWUDO FRQGLWLRQV DQG DVVXPLQJ WKH WKHUPDO GLIIXVLYLW\ WR EH SURSRUWLRQDO WR .A GHULYHG DQ H[SUHVVLRQ IRU .A
PAGE 34
+DQQD f K\SRWKHVL]HG WKDW WKH YHUWLFDO HGG\ GLIIXVLYLW\ GHSHQGV XSRQ WKH SDUDPHWHUV WKDW GHWHUPLQH WKH HQHUJ\ VSHFWUXP RI WKH YHUWLFDO YHORFLW\ IOXFWXDWLRQV 7KH REVHUYDWLRQV RI 5HFRUG DQG &UDPHU f .DLPDO f 6PLWK f DQG RWKHUV ZHUH XVHG WR FDOFXODWH WKH PRPHQWXP PDVV DQG WKHUPDO GLIIXVLYLWLHV +RVLHU f HVWLPDWHG WKH YHUWLFDO GLIIXVLYLWLHV IURP PHDVXUHPHQWV RI DWPRVSKHULF UDGRQ FRQFHQWUDWLRQV 7KH PRGHOLQJ RI FKHPLFDO UHDFWLRQV DQGf¬ LQFRUSRUDWLQJ WKHP LQWR WKH GLIIXVLRQ HTXDWLRQ LV GLVFXVVHG E\ /DPE f +HLQHV DQG 3HWHUV DEf DQG )DEULFN f 7KH VRXUFH KHLJKW XVHG LQ WKH GLVSHUVLRQ PRGHOV LV WKH HIIHFWLYH VWDFN KHLJKW ZKLFK LV WKH VXP RI WKH SK\VLFDO VWDFN KHLJKW DQG WKH SOXPH ULVH 1XPHURXV HPSLULFDO SOXPH ULVH PRGHOV DUH DYDLODEOH LQ OLWHUDWXUH DPRQJ ZKLFK +ROODQGnV HTXDWLRQ 7KRPDV HW DO f LV WKH PRVW ZLGHO\ XVHG 0RVHV DQG 6WRUP f FRPSDUHG REVHUYHG SOXPH ULVHV ZLWK VRPH ZHOONQRZQ IRUPXODV %ULJJVf IURP GLPHQVLRQDO DQDO\VLV DUULYHG DW GLIIHUHQW HTXDn WLRQV IRU WUDQVLWLRQDO VWDEOH DQG QHXWUDO FRQGLWLRQV 8UEDQ DLU SROOXWLRQ PRGHOV 7KH SLRQHHULQJ ZRUN LQ WKH ILHOG ZDV GRQH E\ )UHQNLHO Ef ZKR GHYHORSHG D PRGHO IRU WKH FLW\ RI /RV $QJHOHV /XFDV f LQn WURGXFHG WKH XVH RI WKH *DXVVLDQ SOXPHV 7KLV DSSURDFK ZDV DGRSWHG LQ VHYHUDO PRGHOLQJ VWXGLHV 3RROHU f VWXGLHG WKH QDWXUH RI DWPRVSKHULF GLIIXVLRQ RYHU 6W /RXLV &ODUN f PRGHOHG WKH FLW\ RI &LQFLQQDWL 7XUQHU f GHYHORSHG D PRGHO IRU 1DVKYLOOH DQG
PAGE 35
0LOOHU DQG +RO]ZRUWK f IRU 1HZ
PAGE 36
LLLf 7KH GLIIXVLRQ HTXDWLRQ PRGHOV 7KHVH PRGHOV DUH EDVHG RQ WKH QXPHULFDO VROXWLRQ RI WKH GLIIXVLRQ HTXDWLRQ LQFOXGLQJ WKH FKHPLFDO UHDFWLRQ WHUP 7KHVH PRGHOV DUH IXUWKHU GLYLGHG LQWR Df PRYLQJ FHOO DSSURDFK Ef IL[HG FHOO DSSURDFK ,Q WKH PRYLQJ FHOO DSSURDFK FRQFHQWUDWLRQ FKDQJHV LQ D K\SRWKHWLFDO SDUFHO RI DLU DUH FDOFXODWHG DV LW LV FRQYHFWHG ZLWK WKH ZLQG 7KH ZLQG YHORFLW\ LV DVVXPHG FRQVWDQW ZLWK KHLJKW DQG WKH GLIIXVLRQ LQ WKH KRUL]RQWDO SODQH LV QHJOHFWHG ,Q WKH IL[HG FHOO DSSURDFK WKH DLUVKHG LV GLYLGHG LQWR D WKUHHGLPHQVLRQDO JULG DQG WKH QXPHULFDO VROXWLRQ PD\ XVH D ILQLWH GLIIHUHQFH VFKHPH D YDULDWLRQDO PHWKRG RU D SDUWLFOHLQFHOO 3,&f PHWKRG 7KH 3,& PHWKRG LQYROYHV PRYLQJ WKH PDVV SDUWLFOHV DFFRUGLQJ WR FRQYHFWLRQ DQG GLIIXVLRQ 6WRFKDVWLF 0RGHOV RI 7XUEXOHQFH 2QH RI WKH ILUVW VWRFKDVWLF PRGHOV ZDV GHYHORSHG LQ D 3K' WKHVLV E\ .LUPVH f DQG SUHVHQWHG E\ .LUPVH DQG )DKLHQ f 7KH SUREDELOLW\ GLVWULEXWLRQV RI WKH UDQGRP IRUFHV DFWLQJ RQ WKH IOXLG SDUWLFOHV LQ D UDQGRP IORZ ZHUH FDOFXODWHG ,I WKH PDVV RI D IOXLG SDUWLFOH UHPDLQV FRQVWDQW WKHQ DFFRUGLQJ WR 1HZWRQnV VHFRQG ODZ WKH IRUFH LV SURSRUWLRQDO WR WKH DFFHOHUDWLRQ 7KHUHIRUH LW LV VXIILFLHQW WR FDOFXODWH WKH SUREDELOLW\ GLVWULEXWLRQV RI WKH SDUWLFOH DFFHOHUDWLRQV 7KH SDUWLFOHnV PRYHPHQWV DUH GHVFULEHG E\ WKH /DJUDQJLDQ DFFHOHUDWLRQV DV LW PRYHV ZLWK WKH IORZ 7KH H[SHULn PHQWDO GDWD RI /LHSPDQQ DQG 5RELQVRQ f LQGLFDWHV WKDW WKH
PAGE 37
WXUEXOHQW YHORFLW\ DQG LWV WLPH GHULYDWLYHV KDYH DSSUR[LPDWHO\ QRUPDO GLVWULEXWLRQV 7KLV LQIRUPDWLRQ LV XVHG WR FDOFXODWH WKH VWDWLVWLFDO SDUDPHWHUV DQG WKH SDWKV RI WKH IOXLG SDUWLFOHV DUH VLPXODWHG RQ D GLJLWDO FRPSXWHU 7KLV PRGHO LV XVHG LQ WKH SUHVHQW ZRUN DQG LV GHVFULEHG LQ GHWDLO LQ 6HFWLRQ .LUPVH f VLPXODWHG WKH WXUEXOHQW ILHOG LQ D SLSH XVLQJ WKH PRGHO DQG FDOFXODWHG FRQFHQWUDWLRQ SURILOHV DQG HGG\ GLIIXVLYLW\ FRPSRQHQWV GRZQVWUHDP RI D VRXUFH *LHORZ f XVHG WKH VDPH PRGHO IRU IORZ EHWZHHQ WZR SDUDOOHO SODWHV DQG VKRZHG WKDW WKH PRGHO ZRUNHG IRU KHDW WUDQVIHU DV ZHOO *LHORZ f DOVR VLPXODWHG WKH YHORFLW\ UHFRUG DQG FRUUHODWLRQ FXUYHV DQG WKXV VKRZHG WKDW WKH PRGHO ZDV FDSDEOH RI SUHGLFWLQJ WKH GHWDLOHG VWUXFWXUH .LUPVHrV f ZRUN PXVW EH UHFRJQL]HG DV D SLRQHHULQJ ZRUN LQ WKH QXPHULFDO VLPXODWLRQ RI DQ H[WUHPHO\ FRPSOLFDWHG IORZ VWUXFWXUH RI WXUEXOHQFH ,W VKRXOG DOVR EH QRWHG WKDW WKH .LUPVH)DKLHQ PRGHO LV QRW DQ H[DFW UHSUHVHQWDWLRQ RI D WXUEXOHQW IORZ ILHOG 7KH /DJUDQJLDQ YHORFLW\ DQG DFFHOHUDWLRQ GLVWULEXWLRQV DUH DVVXPHG WR EH RI WKH VDPH QDWXUH $ IOXLG SDUWLFOH LV DVVXPHG WR PDLQWDLQ LWV LGHQWLW\ WKURXJKRXW DQG WKH HQWUDSPHQW RI PDVV E\ WKH HGGLHV LV QHJOHFWHG ,W LV DOVR DVVXPHG WKDW DQ HTXDO DPRXQW RI PDVV RI WKH GLIIXVLQJ VSHFLHV LV FDUULHG E\ DOO WKH HGGLHV ,W LV KRZHYHU TXLWH REYLRXV IURP WKH UHVXOWV WKDW WKHVH HIIHFWV DUH VPDOO HQRXJK WR EH QHJOHFWHG LQ VWXG\LQJ WKH SKHQRPHQRQ RI WXUEXOHQW GLIIXVLRQ 'DYLHV HW DO f DQG *XSWD f QRWHG WKDW 7D\ORUnV f UHSUHVHQWDWLRQ RI SDUWLFOH GLIIXVLRQ E\ GLVFRQWLQXRXV PRYHPHQWV LPSOLHV ]HUR FRUUHODWLRQ EHWZHHQ WKH YHORFLWLHV GXULQJ WZR GLIIHUHQW
PAGE 38
WLPH VWHSV DQG LV WKHUHIRUH DQ LQDGHTXDWH UHSUHVHQWDWLRQ RI WXUEXOHQFH DV ZDV DOVR LQGLFDWHG E\ 7D\ORU f .UDLFKQDQ f VLPXODWHG DQ LVRWURSLF YHORFLW\ ILHOG LQ D WXUEXOHQW IORZ 7KH YHORFLW\ ILHOG ZDV VWRUHG LQ D FRPSXWHU DV D VHW RI )RXULHU FRPSRQHQWV ZKLFK LQ WXUQ ZHUH SLFNHG IURP *DXVVLDQ GLVWULEX WLRQV /DJUDQJLDQ YHORFLW\ FRUUHODWLRQ HGG\ GLIIXVLYLW\ DQG UHVSRQVH IXQFWLRQV ZHUH FDOFXODWHG DQG FRPSDUHG ZLWK WKH UHVXOWV REWDLQHG E\ WKH GLUHFW LQWHUDFWLRQ DSSUR[LPDWLRQ PHWKRG 5REHUWV f 'HDUGRUII DQG 3HVNLQnV f QXPHULFDO PRGHO LQYROYHG REWDLQLQJ WKH HGG\ YHORFLWLHV DW HDFK WLPH VWHS IURP WKH HTXDWLRQ RI PRWLRQ DQG WKHVH ZHUH LQWHJUDWHG WR REWDLQ WKH SDWKV RI WKH SDUWLFOHV 7KH /DJUDQJLDQ FRUUHODWLRQ DQG PHDQ VTXDUH GLVSHUVLRQ ZHUH FDOFXODWHG %XOOLQ DQG 'XNOHU %XOOLQ f XVHG D QXPHULFDO PRGHO VLPLODU WR WKH %URZQLDQ PRWLRQ PRGHO IRU WKH PROHFXODU WUDQVn SRUW DQG XVHG D ZKLWH QRLVH VLJQDO WR UHSUHVHQW WKH PROHFXODU PRWLRQV $ PDMRU GLIIHUHQFH EHWZHHQ WKH .LUPVH)DKLHQ DQG WKH %XOOLQ'XNOHU VWRFKDVWLF PRGHOV LV WKH UHSUHVHQWDWLRQ RI WKH HGG\ YHORFLW\ $Q HGG\ YHORFLW\ ILHOG FRQWLQXRXV LQ VKDSH LV REWDLQHG IURP WKH .LUPVH)DKLHQ PRGHO *LHORZ f DQG LW SRVVHVVHV WKH GHVLUHG SURSHUWLHV :KLOH %XOOLQ DQG 'XNOHU FDUHIXOO\ WDNH LQWR DFFRXQW WKH PROHFXODU DFWLRQ WKH\ UHSUHVHQW WKH HGG\ YHORFLW\ ILHOG E\ D UDWKHU FUXGH DSSUR[LPDWLRQ RI D *DXVVLDQ ZKLWH QRLVH $ ZKLWH QRLVH FRQVLVWV HQWLUHO\ RI XQFRUUHODWHG LPSXOVHV DGMDFHQW WR HDFK RWKHU -HQNLQV DQG :DWWV f ,WV DXWR FRUUHODWLRQ IXQFWLRQ LV D GLUDF GHOWD IXQFWLRQ ZKLFK KDV D ]HUR DXWR FRUUHODWLRQ DW DOO WLPHV W A DQG LV LQILQLWH DW ]HUR WLPH ODSVH 7KH LPSXOVHV DUH UDQGRPO\ GLVWULEXWHG DORQJ WKH
PAGE 39
WLPH D[LV DFFRUGLQJ WR D 3RLVVRQ GLVWULEXWLRQ 6HLQIHOG DQG /DSLGXV f 7KH PRGHO LV VLPLODU WR WKH %URZQLDQ PRWLRQ PRGHO DQG PD\ UHSUHVHQW PROHFXODU WUDQVSRUW DGHTXDWHO\ EXW LV DQ RYHUVLPSOLILFDWLRQ RI WKH FRPSOLFDWHG QDWXUH RI WXUEXOHQFH *ROGVWHLQ f SUHVHQWHG D GLIIHUHQW DQDO\VLV RI GLIIXVLRQ E\ GLVFRQWLQXRXV PRYHPHQWV DOWKRXJK KH FRQVLGHUHG HTXDO WLPH VWHSV DQG D YHORFLW\ RI FRQVWDQW PDJQLWXGH +LV PRGHO LV VLPLODU WR 7D\ORUf¬V f DQDO\VLV RI GLIIXVLRQ E\ GLVFRQWLQXRXV PRYHPHQWV ZLWK WKH GLIIHUHQFH EHLQJ WKDW *ROGVWHLQ LQWURGXFHG D WZR WLPH VWHS FRUUHODWLRQ DIIHFWLQJ WKH VLJQ RI WKH YHORFLW\ 7KLV DVVXPSWLRQ UHVXOWV LQ D WHOHJUDSK HTXDWLRQ LQ FRQWUDVW WR D ILUVW RUGHU GLIIHUHQWLDO HTXDWLRQ IRU WKH RQH VWHS WLPH GHSHQGHQW SURFHVV +RZHYHU LW LV SRVVLEOH WR KDYH D QRQ]HUR FRUUHODWLRQ IRU WKH RQH VWHS GHSHQGHQW SURFHVV LI WKH YHORFLW\ DQG WKH WLPH VWHSV DUH QRW RI FRQVWDQW RUGHU RI PDJQLWXGH %RXUUHW f XVHV D VLPLODU PHWKRG WR WKDW RI *ROGVWHLQ f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
PAGE 40
&+$37(5 7+(25< 7KHRU\ RI 6WRFKDVWLF 3URFHVVHV $ VWRFKDVWLF SURFHVV LV D FROOHFWLRQ RI UDQGRP YDULDEOHV RU D FKDLQ RI HYHQWV WKDW IROORZ D FHUWDLQ SUREDELOLW\ ODZ 7XFNHU .DKQ f 7KH UDQGRP YDULDEOHV FDQ EH FRQVLGHUHG DV WKH YDOXHV RI D VLQJOH UDQGRP YDULDEOH $ UDQGRP YDULDEOH FDQ EH GHILQHG DV WKH QXPHULFDO TXDQWLW\ DVVRFLDWHG ZLWK WKH RXWFRPH RI D UDQGRP SKHQRPHQRQ RU DQ H[SHULPHQWf LQ VXFK D ZD\ WKDW DV WKH YDULRXV HYHQWV RU SRVVLEOH RXWFRPHV RI WKH SKHQRPHQRQ RFFXU WKH UDQGRP YDULDEOH WDNHV RQ GHILQLWH YDOXHV 3DU]HQ )HOOHU f $V H[DPSOHV WKH UDQGRP YDULDEOH FRXOG EH WKH LQVWDQWDQHRXV SRLQW YDOXH RI WKH FRQn FHQWUDWLRQ LQ D UHDFWRU WKH LQVWDQWDQHRXV YDOXH RI WKH YHORFLW\ LQ D WXUEXOHQW IORZ ILHOG RU WKH GLVSODFHPHQW RI D SDUWLFOH GXH WR WXUEXn OHQW GLIIXVLRQ 7KH )RNNHU3ODQFN HTXDWLRQ &RQVLGHU D FRQWLQXRXV VWRFKDVWLF SURFHVV ;BA $V ZH KDYH VHHQ DERYH DW DQ\ WLPH W ;AWf FDQ EH YLHZHG DV WKH QXPHULFDO YDOXH RI D UDQGRP YDULDEOH RU ZH FDQ UHJDUG ;BAWf DV WKH UDQGRP YDULDEOH )RU D FRPSOHWH SUREDELOLVWLF GHVFULSWLRQ RI WKH VWRFKDVWLF SURFHVV DOO WKH PRPHQWV LQILQLWH LQ QXPEHUf DUH UHTXLUHG $OWHUQDWLYHO\ RQH VKRXOG NQRZ DOO WKH MRLQW SUREDELOLW\ GHQVLW\ IXQFWLRQV LH IRU DQ\
PAGE 41
Q S; W ; Wf± ; W f VKRXOG EH NQRZQ 8VLQJ WKH GHIL , ,Q A A Q QLWLRQ RI WKH FRQGLWLRQDO SUREDELOLW\ GHQVLW\ WKLV FDQ EH ZULWWHQ DV S; W ; Wf± ; W f S; Wf± ; W _[ W-S; W f L L LQ Lf± L Qn L L L Q QLO f 6LQFH ZH DUH LQWHUHVWHG P S; W f WKH SUREDELOLW\ GHQVLW\ RI WKH L Q Q VWRFKDVWLF SURFHVV ;A DW WLPH WA ZH FDQ ZULWH (T f DV S; W ; Wf± ; W f L Lf± LQ Q S[L W-[L 9U ;L f¬WLf Q Q S[L WL ; Wf ; WQBf Q f $ 0DUNRY SURFHVV FDQ EH GHILQHG DV D VWRFKDVWLF SURFHVV ZKHUH WKH SUHVHQW VWDWH RI WKH SURFHVV LV FRPSOHWHO\ GHWHUPLQHG E\ LWV SUHn YLRXV VWDWH DQG LV LQGHSHQGHQW RI WKH ZD\ LQ ZKLFK WKH SUHYLRXV VWDWH KDG GHYHORSHG )RU D 0DUNRY SURFHVV WKH FRQGLWLRQDO SUREDELOLW\ GHQVLW\ RQ WKH ULJKW KDQG VLGH RI (T f FDQ EH ZULWWHQ DV 3[L !W _; W ? r WQ f¬ ;L f¬9 S;Lf¬WQ,;L f¬9O Q Q Q Q Q IRU W Wf± W Q f
PAGE 42
(T f LV WKH PDWKHPDWLFDO IRUP RI WKH GHILQLWLRQ RI D 0DUNRY SURFHVV 6HLQIHOG DQG /DSLGXV f 7KH FRQGLWLRQDO SUREn DELOLW\ S; W _; W Wf LV FDOOHG WKH RQH VWHS WUDQVLWLRQ SURE O Q L Qf§ Q Q DELOLW\ 7KH SUREDELOLVWLF GHVFULSWLRQ RI D 0DUNRY SURFHVV GRHV QRW UHTXLUH DQ LQILQLWH QXPEHU RI WKH MRLQW SUREDELOLW\ GHQVLW\ IXQFWLRQV DQG WKH SUREOHP LV UHGXFHG WR ILQGLQJ WKH WUDQVLWLRQ SUREDELOLW\ )URP WKH GHILQLWLRQ RI D 0DUNRY SURFHVV LW LV HDV\ WR VHH WKDW WKH WUDQVLWLRQ SUREDELOLWLHV REH\ WKH &KDSPDQ.ROPRJRURY HTXDWLRQ %KDU XFKD5HLG f 3;W RR S[LW_[IfS[W [ W fG[ R f ZKHUH ;A ;AWnf LV DQ LQWHUPHGLDWH VWDWH RI WKH VWRFKDVWLF SURFHVV DW Wn W! Wn ! W DQG ZKHUH ; DQG W UHSUHVHQW WKH LQLWLDO YDOXHV R L R R RI WKH GLVSODFHPHQW RI WKH SDUWLFOH DQG RI WKH WLPH UHVSHFWLYHO\ 7KH WLPH Wn LV IL[HG LQ WKH LQWHJUDWLRQ RI (TXDWLRQ f 7KH HTXDn WLRQ LPSOLHV WKDW WKH SURFHVV FDQ EH H[SUHVVHG DV JRLQJ IURP WLPH W WR W LQ WZR VWDJHV 1RZ OHW XV FRQVLGHU DQ LQFUHPHQW RI WLPH $W ,I WKH SUHVHQW WLPH LQ WKH &KDSPDQ.ROPRJRURY HTXDWLRQ LV W$Wf LI WKH LQWHUPHGLDWH WLPH LV W DQG LI WKH VWDWH RI WKH VWRFKDVWLF SURFHVV WKH SDUWLFOH GLVSODFHPHQW LV ;A$;A (T f FDQ EH ZULWWHQ DV
PAGE 43
S;W$W@; W f S;W$W_[$;Wf L L R L nOO S;$; ; W fG$;f O nO 2 R f ([SDQGLQJ WKH LQWHJUDQG RQ WKH ULJKW KDQG VLGH RI WKH HTXDWLRQ LQ D 7D\ORU VHULHV LQ WKH [GLUHFWLRQ DERXW WKH SRLQW S;BA$;AW$W M ;AWf fµ S;W_; W f ZH REWDLQ L L R R S;W$W_;$;WfS;$;W_; W f r 2 2 S;$;W$W_[WfS;W M ; W f 2 \ -WXL L;!}e Q L [Q >S;$;W$W_;Wf L L [ Q O S;LW_;L WRf@ R f 6XEVWLWXWLQJ (T f LQWR (T f ZH REWDLQ S;W$W_; W f I S;$;W$W_;Wf3; W_;W fG$; f f« 2 n 2 &2 &2 \ $[ fQ >S; $; W$W_[ WfSm W Qm L L L Q O ;W_; WRf@G$;f R f
PAGE 44
S;W_[ W f S;$;W$W_[WfG$;f RR ; IA7 >S[}WO[L WRf I $;LfQS;L$;W$W_[WfG$;Lf@ Q O ; L f )URP WKH GHILQLWLRQ RI WKH SUREDELOLW\ GHQVLW\ ZH KDYH S;L$;LW$W_;LWfG$;Lf f§22 f 8VLQJ (TXDWLRQ f (TXDWLRQ f FDQ EH UHZULWWHQ DV S[LW$W_[ WRf S[sW_[ WRf eW LE WSnKGK !f«R! R R f§ ; R Q O L $;LfQS;L$;LW$W_;MWfG$;Lf @ f 'LYLGLQJ (TXDWLRQ f E\ $W DQG WDNLQJ WKH OLPLW DV $W r‘ /LP $W /LP $W S;W$W_; W f S;W_; W f L R R R $W ; EU ILU >mLnOK Kr DLU $9Q Q O ; L S;L$;sW$W_;sWfG$;sf f
PAGE 45
7KH OHIW KDQG VLGH RI WKH DERYH HTXDWLRQ LV VLPSO\ WKH GHILQLWLRQ RI WKH SDUWLDO GHULYDWLYH RI WKH WUDQVLWLRQ SUREDELOLW\ ZLWK UHn VSHFW WR WLPH S; W_[ W f R R W /LP $W r S;sW$W_;L WRfS;LW_; W4f R R $W f 7KH Qf§ PRPHQW RI D UDQGRP YDULDEOH ZLWK WKH SUREDELOLW\ GHQVLW\ IXQFWLRQ S; f LV GHILQHG DV 7XFNHU f 22 ( ^[A` ( M [-S; fG; f f f§22 WLO 7KXV WKH UDWH RI FKDQJH RI WKH Qf§ PRPHQW RI WKH LQFUHPHQW $;A FRQGLWLRQHG RQ ;AWf FDQ EH GHILQHG DV D ;Wf QU L L /LP a $;fQS;$;W$W, ;WfG$;f $W $W f 6XEVWLWXWLQJ (TXDWLRQV f DQG f LQWR HTXDWLRQ f WKH IRUZDUG )RNNHU3ODQFN HTXDWLRQ LV REWDLQHG S; W ; W f L n L R R W 0 Q O Q f±Q >D ; Wf S ; ; W f@ Q L [ L R G; L R L f§r ; rr L W ! W f
PAGE 46
7KH ERXQGDU\ FRQGLWLRQV DUH /LP S;W_[ W f ;; f Q Of¬O2 W R 2 f /LP S;W_; W f , BB , L L R ???Ar R (TXDWLRQ f LV DOVR NQRZQ DV WKH IRUZDUG .ROPRJRURY HTXDWLRQ 7KH WKUHHGLPHQVLRQDO )RNNHU3ODQFN HTXDWLRQ LV VLPLODU DQG WKH GHn ULYDWLRQ LV LGHQWLFDO WR WKH RQH GHVFULEHG DERYH H[FHSW IRU (TXDWLRQ f ZKLFK FKDQJHV WR D WKUHHGLPHQVLRQDO 7D\ORU VHULHV DQG KDV DGn GLWLRQDO WHUPV 6WRFKDVWLF GLIIHUHQWLDO HTXDWLRQ .LUPVH f LQ KLV 3K' WKHVLV GHVFULEHV WKH GLVSODFHPHQW RI D SDUWLFOH ; DW WKH HQG RI WKH WLPH VWHS $W DV L P P $W ; ; 9 $W D : f L L L P L P P P P ZKHUH ; LV WKH GLVSODFHPHQW RI WKH SDUWLFOH DW WKH HQG RI WKH P V W Pff§ VWHS DQG DUH WKH YHORFLW\ DQG WKH DFFHOHUDWLRQ UHn VSHFWLYHO\ 7KH GLVSODFHPHQW RI WKH SDUWLFOH DW WKH HQG RI WKH VWHS LV GHWHUPLQHG FRPSOHWHO\ IURP LWV SRVLWLRQ DW WKH HQG RI WKH SUHYLRXV WLPH VWHS 7KH VLQJOH PRVW UHFHQW YDOXH RI WKH GLVSODFHPHQW FRQWDLQV DV PXFK LQIRUPDWLRQ DERXW HYDOXDWLQJ ; DW W DV GRHV WKH L P P
PAGE 47
HQWLUH SDVW KLVWRU\ RI WKH SDUWLFOH 7KHUHIRUH WKH GLVSODFHPHQW RI D SDUWLFOH LQ WKH .LUPVH)DKLHQ PRGHO LV D 0DUNRY SURFHVV $ VWRn FKDVWLF GLIIHUHQWLDO HTXDWLRQ IRU WKH PRGHO LV GHULYHG LQ WKLV VHFn WLRQ FRQVLGHULQJ WKH GLVSODFHPHQW RI D SDUWLFOH DV D 0DUNRY SURFHVV $ JHQHUDO VWRFKDVWLF GLIIHUHQWLDO HTXDWLRQ InRU D 0DUNRY SURFHVV FDQ EH ZULWWHQ DV 6WUDWRQRYLFK f G; L GW :r} LI rfµr! f ZKHUH ;BA LV D VWRFKDVWLF SURFHVV ZLWK 0DUNRYLDQ SURSHUWLHV IA;BAWf LV WKH GHWHUPLQLVWLF IXQFWLRQ DQG JBA;AWf LV WKH UDQGRP IXQFWLRQ RI WKH SURFHVV 7KH ILUVW WZR PRPHQWV RI WKH GLVSODFHPHQW $;A FDQ EH FDOFXn ODWHG DV &2 b ‘ D6R 67 0LfS;$;W$W_[LWfG$;Of f§&2 ILmOm f DQG ; L G( GW f
PAGE 48
7KH GHWHUPLQLVWLF IXQFWLRQ I LV WKHQ WKH GHWHUPLQLVWLF L YHORFLW\ LQ WKH Lf§ GLUHFWLRQ LQ D WXUEXOHQW IORZ DQG WKH UDQGRP IXQFWLRQ GJBGW LV WKH UDQGRP YHORFLW\ FRQWULEXWLRQ LQ WKH Lf§ GLUHFWLRQ 'LIIHUHQWLDWLQJ (T f ZLWK UHVSHFW WR WLPH DQG XVLQJ YHFWRU QRWDWLRQ ZH KDYH f ZKHUH ; LV WKH GLVSODFHPHQW YHFWRU RI D SDUWLFOH LQ VSDFH G ;GW n LV LWV VHFRQG GHULYDWLYH ZLWK UHVSHFW WR WLPH DV LW PRYHV LQ WKH VSDFH DQG LV WKHUHIRUH WKH /DJUDQJLDQ DFFHOHUDWLRQ YHFWRU 7KH GHULYDWLYH GIGW UHSUHVHQWV WKH PHDQ /DJUDQJLDQ DFFHOHUDWLRQ RQ WKH SDUWLFOH DQG G MrGW f LV WKH IOXFWXDWLQJ /DJUDQJLDQ DFFHOHUDWLRQ 7KHQ f Dn f ZKHUH D DQG Dn DUH WKH PHDQ DQG WKH IOXFWXDWLQJ FRPSRQHQWV RI WKH /DJUDQJLDQ DFFHOHUDWLRQ YHFWRU UHVSHFWLYHO\ (TXDWLRQ f FDQ QRZ EH ZULWWHQ DV
PAGE 49
G; D D Dn GW f 7KH .LUPVH)DKLHQ PRGHO XVHV HPSLULFDO SUREDELOLW\ GLVWULn EXWLRQV IRU HDFK RI WKH FRPSRQHQWV RI WKH /DJUDQJLDQ DFFHOHUDWLRQ DQG VLPXODWHV WKH SDWK RI IOXLG SDUWLFOHV RQ D GLJLWDO FRPSXWHU 7KH PRGHO LV GLVFXVVHG LQ GHWDLO LQ 6HFWLRQ 7KH LQWHJUDWHG IRUP RI (T f LV WKH VWRFKDVWLF GLIIHUn HQWLDO IRU D PRGHO FRQVLGHULQJ WKH GLVSODFHPHQW DV D 0DUNRY SURFHVV ,QWHJUDWLQJ (T f WKH VWRFKDVWLF GLIIHUHQWLDO HTXDWLRQ LV H[n SUHVVHG LQ WHUPV RI WKH /DJUDQJLDQ DFFHOHUDWLRQV DQG LQ WKH /DJUDQJLDQ FRRUGLQDWH V\VWHP DV /DJ W R W D GW 9 f§ R W R Dn GW f ZKHUH LV WKH WXUEXOHQW YHORFLW\ YHFWRU SRVVHVVHG E\ WKH SDUWLFOH DW WKH EHJLQQLQJ RI WKH VLPXODWLRQ 7KH DERYH HTXDWLRQ LQ WKH (XOHU LDQ IUDPH RI UHIHUHQFH LV VLPSO\ WKH IROORZLQJ G; GW (XO 9 W R W D GW 9n W R ZKHUH 9 LV WKH PHDQ YHORFLW\ YHFWRU Dn GW f
PAGE 50
&RPSDULVRQ RI WKH )RNNHU3ODQFN HTXDWLRQ ZLWK WKH GLIIXVLRQ HTXDWLRQ 7KH )RNNHU3ODQFN HTXDWLRQ IRU D IRU Q ! UHGXFHV WR Q f§ L O( W ;W L r L R R ; L DL Sf L r 3f L f ZKHUH LV WKH SDUWLFOH GLVSODFHPHQW LQ WKH Lf§ GLUHFWLRQ 7KH ILUVW PRPHQW LQ WKH Lf§f§ GLUHFWLRQ IURP (TXDWLRQV f DQG L f WKHQ LV D I;Wf 9 I D LL L L GW f 7KH ILUVW PRPHQW LQ WKH .LUPVH)DKLHQ PRGHO D LV WKHUHIRUH L WKH GHWHUPLQLVWLF YHORFLW\ ZKLFK LV WKH VXP RI WKH DYHUDJH YHORFLW\ LQ WKH IORZ ILHOG DQG DQRWKHU WHUP DULVLQJ GXH WR WKH PHDQ /DJUDQJLDQ DFFHOHUDWLRQ 7KH SK\VLFDO PHDQLQJ RI DLV HDV\ WR VHH IURP (TXDWLRQ L f ZKLFK FDQ EH DOWHUQDWLYHO\ ZULWWHQ DV B G U Ds n GW L` f ZKHUH G;BA LV WKH H[SHFWHG YDOXH RI G; f FRQGLWLRQHG DW ;AWf
PAGE 51
*, 7D\ORU f H[SUHVVHV GLIIXVLYLW\ DV WKH WLPH GHULYDWLYH RI WKH YDULDQFH RI GLVSHUVLRQ NW G GW rs f 7KHUHIRUH LV D SRLQW HGG\ PDVV GLIIXVLYLW\ ZKLFK LV L D IXQFWLRQ RI WKH SDUWLFOH SRVLWLRQ DQG WLPH 7KH GLIIXVLYLW\ L LV EDVHG XSRQ WKH /DJUDQJLDQ VWDWLVWLFV RI WKH SDUWLFOHV /DPE DQG 6HLQIHOG f DQG PD\ WKHUHIRUH EH FDOOHG DV WKH /DJUDQJLDQ GLIIXn VLYLW\ L 7KH )RNNHU3ODQFN HTXDWLRQ FDQ EH ZULWWHQ LQ WHUPV RI WKH FRQn FHQWUDWLRQ LQVWHDG RI WKH WUDQVLWLRQ SUREDELOLW\ 7KH GLVSODFHPHQW FRUUHVSRQGV WR WKH FRRUGLQDWH ZKHQ ZH JR IURP SUREDELOLW\ VSDFH WR WLPH DYHUDJHG FRQFHQWUDWLRQ VSDFH 7KHQ f f 6XEVWLWXWLQJ (TXDWLRQV f DQG f LQWR (TXDWLRQ f DQG XVLQJ 3;AWf FWf f¯O 9L I Ds A[Lf¬ L A Wf GW ZH KDYH
PAGE 52
F [ Wf W [ L >9L[AWf Ds [Wf GW@ F [Wf [sWf F[sWf@ [ L L f 7KH WLPHDYHUDJHG GLIIXVLRQ HTXDWLRQ LQFOXGLQJ WKH FRQYHFn WLYH IORZ WHUP LV F[sWf W  >.L[Lf¬W` VO FYWf@ f ; >9 [ Wf F[ Wf @ R[ OL O 7KH (XOHULDQ GLIIXVLYLW\ . FDQ EH UHODWHG WR WKH /DJUDQJLDQ GLIIX VLYLW\ E\ HTXDWLQJ (TXDWLRQ f DQG f DQG VROYLQJ IRU L . L D L GW 4F6[Af m/ &f f ZKHUH W4 UHSUHVHQWV WKH FRQGLWLRQV DW WKH VRXUFH 7KH WZR GLIIX VLYLWLHV DUH HTXDO WR HDFK RWKHU LI KRPRJHQHRXV FRQGLWLRQV H[LVW DQG LI WKH /DJUDQJLDQ DFFHOHUDWLRQV DUH HTXDO WR ]HUR
PAGE 53
7KH /DJUDQJLDQ GLIIXVLYLW\ LQ JHQHUDO VKRXOG EH H[SHFWHG WR EH D IXQFWLRQ RI WKH VRXUFH DQG RI WKH LQLWLDO WLPH DV ZHOO DV RI WKH WLPH RI GLIIXVLRQ $ TXHVWLRQ WKHQ DULVHV DV WR ZKHWKHU WKH (XOHULDQ GLIIXVLYLW\ . VKRXOG DOVR EH D IXQFWLRQ RI WKH VRXUFH L 0ROHFXODU GLIIXVLYLW\ LV D SURSHUW\ RI WKH PHGLXP DQG LI DQ HGG\ GLIn IXVLYLW\ LV WR EH GHILQHG LQ DQ DQDORJRXV PDQQHU LW VKRXOG EH D SURSHUW\ RI WKH IORZ ILHOG RQO\ UDWKHU WKDQ D TXDQWLW\ DVVRFLDWHG ZLWK WKH VRXUFH DQG WKH WLPH RI GLIIXVLRQ 7KH UDWH RI FKDQJH RI WKH PRPHQWV RI WKH WKLUG DQG KLJKHU RUGHU D n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
PAGE 54
VWUXFWXUH )RU IUHH VKHDU VWHDG\ VWDWH IORZV LW LV DVVXPHG WKDW WKH VWUXFWXUH RI WKH IORZ LV WKH VDPH DW GLIIHUHQW GRZQVWUHDP GLVn WDQFHV 0DWKHPDWLFDOO\ WKLV DVVXPSWLRQ LV H[SUHVVHG DV XWf X ZWf )Qf f ZKHUH XWf LV WKH LQVWDQWDQHRXV YHORFLW\ RI WKH SDUWLFOH X LV LWV PHDQ YHORFLW\ ZWf LV D YHORFLW\ VFDOH RI WKH SDUWLFOH PRWLRQ DW WLPH W DQG WKH LQGHSHQGHQW YDULDEOH U_ LV SURSRUWLRQDO WR WKH UDWLR RI WKH WLPH GW WR 7 WKH WLPH VFDOH RI WKH SDUWLFOH PRWLRQ GQ D GW 7Wf f 7KH IXQFWLRQ ) LV VWDWLRQDU\ DQG UDQGRP 7KH VWDWLVWLFDO PHDQ YHORFLW\ RI DOO WKH SDUWLFOHV LV WKHQ REWDLQHG E\ DYHUDJLQJ (T f RYHU DOO WKH SDUWLFOHV 7KXV X )Qf ZWf f ZKHUH ; LV WKH GLVSODFHPHQW RI WKH SDUWLFOH 7KH IXQFWLRQ )U_f LV REWDLQHG E\ DVVXPLQJ WKDW WKH OHQJWK DQG WKH YHORFLW\ VFDOHV RI WKH SDUWLFOH PRWLRQ DUH JLYHQ UHVSHFWLYHO\ E\ eWf D >;Wf@ 3 f
PAGE 55
DQG ZWf D >;Wf@f« f 7KH WLPH VFDOH RI WKH PRWLRQ FDQ EH REWDLQHG E\ WDNLQJ WKH UDWLR RI WKH OHQJWK WR WKH YHORFLW\ VFDOH UWf >;Wf@ 3LS f %DWFKHORU f ODWHU VXJJHVWHG WKDW RQH VKRXOG EH DEOH WR REWDLQ WKH VDPH UHVXOWV IRU D JURXQG OHYHO VWHDG\ VWDWH VRXUFH LQ WKH DWPRVn SKHULF JURXQG OD\HU +H GHYHORSHG WKLV LGHD LQ DQ XQSXEOLVKHG ZRUN ZKLFK LV GLVFXVVHG E\ (OOLVRQ f &HUPDN f DQG &VDQDG\ f ,W ZDV DVVXPHG WKDW WKH (XOHULDQ SURSHUWLHV RI WKH IORZ VWUXFWXUH GHn SHQG RQO\ XSRQ WKH IULFWLRQ YHORFLW\ LQ D QHXWUDOO\ VWUDWLILHG OD\HU DQG WKDW WKH URXJKQHVV OHQJWK DIIHFWV WKH PHDQ YHORFLW\ RI WKH IORZ RQO\ 7KH VLPLODULW\ K\SRWKHVLV IRU HOHYDWHG VRXUFHV LV H[SUHVVHG E\ &HUPDN f DV IROORZV )RU D PDUNHG SDUWLFOH ZKLFK LV DW ] + ZKHQ W WKH VWDWLVWLFDO SURSHUWLHV RI SDUWLFOH PRWLRQ DW WLPH W GHSHQG RQO\ XSRQ Xr DQG WW ZKHUH W LV RI RUGHU r9Xr RU ODUJHU ZKHUH W LV D YLUWXDO WLPH RULn JLQ ZLWK PDJQLWXGH RI RUGHU +X SDJH f 7KH PHDQ ORQJLWXGLQDO DQG WKH YHUWLFDO SRVLWLRQV RI WKH SDUn WLFOHV FDQ EH H[SUHVVHG DV G= GW f
PAGE 56
DQG G; GW Xr rWU! IRU W } + X 7KH FRQFHQWUDWLRQ LV DVVXPHG WR EH RI WKH IRUP P F D [ f f DQG WKH SOXPH ZLGWK DW WKH JURXQG LV DVVXPHG WR EH RI WKH IRUP < D [ f %DVHG RQ WKH WZR DERYH PHQWLRQHG UHODWLRQVKLSV WKH H[SRQHQW IRU WKH SOXPH ZLGWK DW WKH JURXQG LV IRXQG WR EH f +RZHYHU %DWFKHORUnV K\SRWKHVLV f DQG &HUPDNnV f GHYHORSPHQWV DUH IRU D WLPH RI UHOHDVHV ODUJHU WKDQ +XA )RU D VWDFN KHLJKW RI P DQG D IULFWLRQ YHORFLW\ RI PSV W VKRXOG EH ODUJHU WKDQ VHFRQGV $W D YHORFLW\ RI PSV RQH PXVW EH DW
PAGE 57
OHDVW NP DZD\ IURP WKH VRXUFH &HUPDN f FRPSDUHG WKH WKHRU\ ZLWK VRPH DWPRVSKHULF GLIIXVLRQ H[SHULPHQWV DQG ZLWK VRPH ZLQG WXQn QHO GDWD +H FRQFOXGHG WKDW WKH UDWLR RI VRXUFH KHLJKW WR URXJKQHVV OHQJWK LV RI PDMRU LPSRUWDQFH LQ DWPRVSKHULF GLVSHUVLRQ FDOFXn ODWLRQV 7KH JURZWK ODZ IRU D 0DUNRY SURFHVV 7KH /DJUDQJLDQ VLPLODULW\ WKHRU\ RI %DWFKHORU LV PRGLILHG IRU VPDOO WLPHV EDVHG RQ WKH FRUUHODWLRQ FRHIILFLHQW UHVXOWV )RU D 0DUNRY SURFHVV WKH FRUUHODWLRQ FRHIILFLHQW LV RI WKH H[SRQHQWLDO GHFD\ W\SH &VDQDG\ f 5M7f H 7 f ZKHUH 7 LV WKH LQWHJUDO /DJUDQJLDQ WLPH VFDOH ,I = UHSUHVHQWV WKH GLVSODFHPHQW RI D SDUWLFOH LQ ]GLUHFWLRQ ZH FDQ ZULWH 7D\ORU f Wn 7 ZKHUH Df° LV WKH YDULDQFH RI WKH GLVSHUVLRQ LQ WKH ]GLUHFWLRQ ] )RU QHXWUDO FRQGLWLRQV 3DQRIVN\ DQG 0F&RUPLFN &UDPHU DQG 5HFRUG f G7 GWn f G9 GW M G 2 ] GW Z Z D X ‘N , : 9r f
PAGE 58
ZKHUH LV D XQLYHUVDO FRQVWDQW /HW [ \ DQG ] EH WKH PHDQ SRVLWLRQV RI WKH SDUWLFOHV UHn OHDVHG IURP WKH VRXUFH 6LQFH WKH UDWH RI FKDQJH RI ] LV YHU\ VPDOO FRPSDUHG WR WKH UDWH RI FKDQJH RI [ DQG VLQFH D[LDO GLIIXVLRQ FDQ G[ G[ EH QHJOHFWHG RQH FDQ ZULWH A DV WKH PHDQ YHORFLW\ RI WKH IOXLG DW ] ] &HUPDN f )RU D ORJDULWKPLF YHORFLW\ SURILOH G[ GW f 6XEVWLWXWLQJ (TXDWLRQV f DQG f LQWR (T f ZH KDYH  E  H $/ G[ G[n f ,QWHJUDWLQJ (T f WZLFH ZLWK UHVSHFW WR [ ZH REWDLQ WKH IROORZLQJ H[SUHVVLRQ D ] E A >,Q f§ f@ ] H f D ] ; $W f ZKHUH EO. E ,Q ILf f
PAGE 59
(T f FDQ EH ZULWWHQ DV 2 ] E $ f 6LQFH ] FKDQJHV YHU\ OLWWOH ZLWK [ LW PD\ EH UHSODFHG E\ WKH VWDFN KHLJKW + (T f FDQ EH QRQGLPHQVLRQDOL]HG E\ + WR REWDLQ DQ r H[SUHVVLRQ IRU WKH GLVSHUVLRQ FRHIILFLHQW & A]+ ] r D ] r ; f ZKHUH ? r E Z E N f f ([SDQGLQJ (T f ZH REWDLQ r E D f§ ] 67 ; ; $O $ [ r$O f 7KH LQILQLWH VHULHV DERYH FDQ EH DSSUR[LPDWHG E\
PAGE 60
RU r 2 ] E f f 7KH JURZWK ODZ IRU D 0DUNRY SURFHVV VKRXOG WKHUHIRUH EH JLYHQ E\ (T f ,W LV LPSRUWDQW WR VHH LI WKLV EHKDYLRU LV REWDLQHG IRU WKH HPSLULFDO FRUUHODWLRQ FRHIILFLHQW FXUYHV DV ZHOO 6LQFH WKH /DJUDQJLDQ FRUUHODWLRQ FRHIILFLHQWV DUH DVVXPHG WR EH VLPLODU WR WKH (XOHULDQ FRUUHODWLRQ FRHIILFLHQWV 6HF f LW LV SRVVLEOH WR REWDLQ WKH VKDSHV RI YV W FXUYHV E\ LQWHJUDWLQJ (T f WZLFH U W Wf¬ D D ] 5JWf G[ GWn , W f 7KH H[SHULPHQWDO GDWD RI 3UDLULH *UDVV ZDV XVHG IRU WKH FRUn UHODWLRQ FRHIILFLHQW WR REWDLQ WKH ULJKW KDQG VLGH RI (T f ZKLFK ZH ZLOO FDOO A 7KH UHVXOWLQJ FXUYHV DUH ILWWHG WR WKH IRUP W EW f
PAGE 61
7KLV FXUYH ILW ZDV FDUULHG RXW IRU DOO WKH VWDELOLW\ FRQGLn WLRQV 7KH UHVXOWV IRU RQH W\SLFDO QHXWUDO 5XQ f RQH XQVWDEOH 5XQ f DQG RQH VWDEOH 5XQ f FRQGLWLRQV DUH VKRZQ LQ )LJXUHV f WR f 7KH VKDSH RI WKH JURZWK ODZ REWDLQHG IURP WKH FRUn UHODWLRQ FXUYHV LV LQ H[FHOOHQW DJUHHPHQW ZLWK (T f ZKLFK LV REWDLQHG E\ DVVXPLQJ WKDW WKH GLIIXVLRQ SURFHVV LV D 0DUNRY SURFHVV DQG FRQVHTXHQWO\ WKH FRUUHODWLRQ FXUYHV DUH RI DQ H[SRQHQWLDO GHFD\ W\SH 7KH VDPH IRUP RI WKH JURZWK ODZ LV H[SHFWHG WR EH YDOLG IRU WKH GLVSHUVLRQ LQ WKH \GLUHFWLRQ DV ZHOO 7KXV r D \ r ; E A f $OWKRXJK WKH HIIHFW RI VWDELOLW\ FRQGLWLRQV RQ WKH IRUP RI WKH JURZWK ODZ LV DVVXPHG QHJOLJLEOH WKH SDUDPHWHUV DUH IXQFWLRQV RI VWDELOLW\ FRQGLWLRQV DQG WKH\ VKRXOG EH FRUUHODWHG ZLWK WKH SURSHU VWDELOLW\ SDUDPHWHUV
PAGE 62
)LJXUH &RPSDULVRQ RI , FXUYH FDOFXODWHG IURP WKH FRUUHODWLRQ FRHIILFLHQW GDWD ZLWK WKH 0DUNRY SURFHVV FXUYH ILW IRU W\SLFDO QHXWUDO FRQGLWLRQV
PAGE 63
VHF )LJXUH &RPSDULVRQ RI , FXUYH FDOFXODWHG IURP WKH FRUUHODWLRQ FRHIILFLHQW GDWD ZLWK WKH 0DUNRY SURFHVV FXUYH ILW IRU W\SLFDO XQVWDEOH FRQGLWLRQV
PAGE 64
VHF )LJXUH &RPSDULVRQ RI , FXUYH FDOFXODWHG IURP WKH FRUUHODWLRQ FRHIILFLHQW GDWD ZLWK WKH 0DUNRY SURFHVV FXUYH ILW IRU W\SLFDO VWDEOH FRQGLWLRQV
PAGE 65
&+$37(5 7+( 672&+$67,& 02'(/ 'HVFULSWLRQ RI WKH 0RGHO 'LIIXVLRQ IURP D VWHDG\ VWDWH FRQWLQXRXV SRLQW VRXUFH LV PRGHOHG LQ WKLV ZRUN )RU WKH DSSOLFDWLRQ RI WKH PRGHO XVHG LQ WKLV ZRUN WKH RULJLQ LV DW WKH JURXQG OHYHO GLUHFWO\ EHORZ WKH VRXUFH WKH [D[LV H[WHQGV KRUL]RQWDOO\ LQ WKH GLUHFWLRQ RI WKH PHDQ ZLQG WKH \D[LV LV LQ WKH KRUL]RQWDO SODQH SHUSHQGLFXODU WR WKH [D[LV FURVV ZLQGf DQG WKH ]D[LV LV LQ WKH YHUWLFDO GLUHFWLRQ 7KH FRQn WLQXRXV HPLVVLRQ RI SROOXWDQWV IURP D VWDFN LV DVVXPHG WR EH FRPSRVHG RI D ILQLWH ODUJH QXPEHU RI PDVV SRFNHWV RI WKH RUGHU RI f (DFK RI WKHVH PDVV SRFNHWV LV UHIHUUHG WR DV D SDUWLFOH (DFK SDUWLFOH LV UHOHDVHG IURP WKH VRXUFH ZLWK D VWDWLVWLFDOO\ UDQGRP YHORFLW\ 7KH PRWLRQ RI WKH SDUWLFOH LQ DFFRUGDQFH ZLWK D VWRFKDVWLF HTXDWLRQ LV VLPXODWHG E\ D 0RQWH &DUOR PHWKRG XQWLO LW H[LWV WKH UHJLRQ RI LQWHUHVW 6WDWLVWLFDO DQDO\VLV RI D YHORFLW\ ILHOG LQ D VWHDG\ VWDWH WXUEXOHQW IORZ LQGLFDWHV WKDW WKH YHORFLW\ KDV DQ DSSUR[LPDWH PDUn JLQDO *DXVVLDQ GLVWULEXWLRQ ZLWK FRQVWDQW YDULDQFH DQG WKDW WKH WLPH GHULYDWLYHV RI WKH YHORFLWLHV DUH DOVR FRQWLQXRXV DQG DOVR KDYH DQ DSSUR[LPDWH PDUJLQDO QRUPDO GLVWULEXWLRQ 6LPPRQV DQG 6DOWHU 7RZQVHQG f *DXVVLDQ GLVWULEXWLRQV ZLWK PHDVXUHG RU FRUUHODWHG
PAGE 66
YDULDQFHV DUH XVHG WR VLPXODWH WKH YHORFLWLHV RI WKH IOXLG SDUWLFOHV 7KH YDULDQFHV RI WKH DFFHOHUDWLRQV DUH FDOFXODWHG IURP LQIRUPDWLRQ UHJDUGLQJ WKH YHORFLW\ FRUUHODWLRQV DQG WKH UHODWLRQV EHWZHHQ WKH /DJUDQJLDQ DQG (XOHULDQ FRUUHODWLRQV )RU D GLJLWDO VLPXODWLRQ LW LV QHFHVVDU\ WR GLVFUHWL]H WKH SUREDELOLW\ GLVWULEXWLRQ FXUYHV 6HH $SSHQGL[ $f 7KH SDUWLFOHV DUH PRYHG LQ ILQLWH WLPH VWHSV EXW WKH VLPXODWHG LQVWDQWDQHRXV YHORFLW\ ILHOG LV FRQWLQXRXV ,Q WKH SUHYLRXV FKDSWHU DQ HTXDWLRQ IRU WKH LQVWDQWDQHRXV DFFHOHUDWLRQ ZDV GHULYHG (T f IURP WKH JHQHUDO VWRFKDVWLF HTXDWLRQ IRU D 0DUNRY SURFHVV 7KH GLVSODFHPHQW $; RI f¯P WWO WLO WKH SDUWLFOH LQ WKH Lf§ GLUHFWLRQ GXULQJ WKH Pf§ WLPH VWHS $W LV REn WDLQHG E\ LQWHJUDWLQJ (T f WZLFH LQ WKH Lf§ GLUHFWLRQ RYHU $W P 7KHQ P D L GW ,I DGW L f ZKHUH 9 LV WKH WRWDO YHORFLW\ RI WKH SDUWLFOH DW WKH EHJLQQLQJ PO RI WKH WLPH VWHS ,I WKH /DJUDQJLDQ DFFHOHUDWLRQV DUH DVVXPHG FRQn VWDQW GXULQJ WKH WLPH VWHS $W $W $W $; 9 $W P P P D L P D P P P f 7KH WRWDO YHORFLW\ 9 6QO LV WKH VXP RI WKH PHDQ YHORFLW\ 9 PO DQG WKH
PAGE 67
IOXFWXDWLQJ YHORFLW\ 7KH PRGHO LV FDSDEOH RI XVLQJ DQ\ JLYHQ IRUP RI WKH PHDQ YHORFLW\ 9 [\]f 7KH PHDQ /DJUDQJLDQ DFFHOHUD 6QO WLRQV DUH REWDLQHG IURP WKH HTXDWLRQ RI PRWLRQ E\ HVWLPDWLQJ GLIn IHUHQW WHUPV IURP WKH PHDVXUHG PHWHRURORJLFDO SDUDPHWHUV 7KH IOXFn WXDWLQJ DFFHOHUDWLRQV DUH REWDLQHG UDQGRPO\ IURP *DXVVLDQ GLVWULEXn WLRQV 6HH $SSHQGL[ $f 7KH WLPH VWHSV DUH UDQGRP DQG ZLWKRXW DQ\ PHPRU\ DQG WKXV FDQ EH REWDLQHG IURP DQ H[SRQHQWLDOO\ GLVWULEXWHG SRSXODWLRQ .LUPVH f 7KH PHDQ WLPH VWHS FDQ EH ORRNHG XSRQ DV D WLPH VFDOH RI WXUEXOHQFH 6LQFH LW LV DVVXPHG WKDW QR LQWHUDFWLRQ WDNHV SODFH EHWZHHQ LQGLYLGXDO SDUWLFOHV WKH\ DUH UHOHDVHG IURP WKH VRXUFH LQ VXFFHVVLRQ 7KH SRVLWLRQ RI WKH SDUWLFOH DW D IL[HG LQWHUYDO RI WLPH WJ LV VWRUHG LQ WKH FRPSXWHU WR JHW WKH FRQFHQWUDWLRQ YDOXHV 7KH FRQFHQWUDWLRQ LV FDOFXODWHG IURP WKH SDUWLFOH VDPSOLQJ DFFRUGLQJ WR WKH IROORZLQJ HTXDWLRQ 1 f±W f FHOO V FHOO f ZKHUH 4 1 3 1 FHOO FHOO VRXUFH VWUHQJWK 07 QXPEHU RI SDUWLFOHV UHOHDVHG QXPEHU RI SDUWLFOHV SRVLWLRQV VDPSOLQJVf LQ D FHOO VDPSOLQJ WLPH 7 YROXPH RI D FHOO 0
PAGE 68
,W FDQ EH VKRZQ WKDW LV WKH VRXUFH VWUHQJWK UHSUHVHQWHG E\ HDFK SDUWLFOH LV WKH PDVV RI WKH SROOXWDQW UHSUHVHQWHG E\ HDFK VDPSOLQJ LV WKH PDVV RI WKH SROOXWDQW JLYHQ E\ DOO WKH VDPSOLQJV LQ D FHOO 7RWDO UHIOHFWLRQ LV DVVXPHG DW WKH JURXQG LH ZKHQ WKH SDUWLFOHV KLW WKH JURXQG WKH\ DUH UHIOHFWHG LQ D SHUIHFWO\ HODVWLF PDQQHU 7KH PROHFXODU PRWLRQ PD\ EH QHJOHFWHG LQ WKH DWPRVSKHUH EHn FDXVH RI YHU\ ODUJH WLPH DQG OHQJWK VFDOHV RI WXUEXOHQFH 0RQLQ DQG
PAGE 69
)RU WKH VLPXODWLRQ RI (T f WKH IROORZLQJ SDUDPHWHUV DUH QHHGHG WKH VWDQGDUG GHYLDWLRQV RI HDFK RI WKH FRPSRQHQWV RI WKH IOXFWXDWLQJ YHORFLW\ WXUEXOHQFH GHQVLW\f WKH PHDQ /DJUDQJLDQ DFn FHOHUDWLRQ SURILOH WKH VWDQGDUG GHYLDWLRQV RI WKH IOXFWXDWLQJ FRPn SRQHQWV RI WKH /DJUDQJLDQ DFFHOHUDWLRQV DQG WKH PHDQ YDOXHV RI WKH WLPH VWHSV 6LQFH RQO\ WKH (XOHULDQ SURSHUWLHV DUH XVXDOO\ PHDVXUHG DQG UHSRUWHG /DJUDQJLDQ SURSHUWLHV DUH REWDLQHG WKURXJK WKH XVH RI WKH IDFWRU % 6 $H$O f ? ZKHUH $A DQG $A DUH WKH (XOHULDQ DQG /DJUDQJLDQ PDFURVFRSLF OHQJWK VFDOHV UHVSHFWLYHO\ .LUPVH f DQG *LHORZ f XVHG D YDOXH RI % +D\ DQG 3DVTXLOO f SURSRVH D FRQVWDQW UDWLR RI WKH /DJUDQJLDQ DQG (XOHULDQ WLPH VFDOHV LH 5MHR AWf f ZKHUH DQG WLRQ FRHIILFLHQWV UHVSHFWLYHO\ 7KHLU HVWLPDWH RI YDULHG IURP WR $QJHO¯ QW DO f UHSRUWHG IURP WR +RZHYHU WKH\ HVWLPDWHG IURP WKH UDWLRV RI WKH IUHTXHQFLHV DW ZKLFK WKH PD[LPDV DUH REVHUYHG IRU WKH /DJUDQJLDQ DQG WKH (XOHULDQ YHORFLW\ VSHFWUD 7KHVH DUH SRLQW YDOXHV RI DQG GR QRW UHSUHVHQW WKH YDOXHV 5J DUH WKH /DJUDQJLDQ DQG (XOHULDQ YHORFLW\ FµUUHOD
PAGE 70
WKDW PLJKW EH VDWLVIDFWRU\ RYHU WKH FRPSOHWH UDQJH RI IUHTXHQF\ IRU VSHFWUDf RU WLPH IRU FRUUHODWLRQVf ,VOLW]HU DQG 'XPEDXOG f HVWLPDWHG IRU DOO WKHLU UXQV 6LQFH WKHLU YDOXH LV LQWHUPHGLDWH EHWZHHQ WKRVH RI +D\ DQG 3DVTXLOO WKLV YDOXH LV XVHG WKURXJKRXW WKLV ZRUN 7KH IDFWRU % LV FDOFXODWHG IURP DV IROORZV /HW 7 DQG 7 EH WKH /DJUDQJLDQ DQG (XOHULDQ PDFURVFRSLF / &M WLPH VFDOHV UHVSHFWLYHO\ 7KHQ 22 f DQG 22 GW f ([SUHVVLQJ 7 LQ WHUPV RI 7 / ( &2 22 f 8VLQJ (T f 22 GW 7 ( f
PAGE 71
H WOWH f DQG % $H$O f 8VLQJ ZHOO NQRZQ UHODWLRQV EHWZHHQ WKH VSDFH DQG WLPH PDFURVFDOHV +LQ]H f $ ( DQG WKHUHIRUH % R Y ; 7 / f f f ZKHUH 2 LV WKH LQWHQVLW\ RI WXUEXOHQFH LQ WKH Lf§ GLUHFWLRQ L %\ WDNLQJ WKH )RXULHU WUDQVIRUP RI (T f WKH /DJUDQJLDQ DQG (XOHULDQ HQHUJ\ VSHFWUD FDQ EH FRUUHODWHG DV IROORZV +D\ DQG 3DVTXLOO f )/Qf % 9Qf f
PAGE 72
ZKHUH ) DQG ) DUH WKH /DJUDQJLDQ DQG (XOHULDQ HQHUJ\ VSHFWUD UHVSHFWLYHO\ DQG Q LV WKH IUHTXHQF\ (T f ZKHQ LQWHJUDWHG RYHU WKH HQWLUH IUHTXHQF\ UDQJH JLYHV WKH HTXDOLW\ EHWZHHQ WKH /DJUDQJLDQ DQG (XOHULDQ WXUEXOHQFH HQHUJLHV %XW %XOOLQ DQG 'XNOHU f XVHG LQ WKH IROORZLQJ HTXDWLRQ ZKLFK LV DQ DSSUR[LPDWLRQ RI (T f )/Qf )eQf f $ YDOXH RI ZRXOG LPSO\ WKDW WKH /DJUDQJLDQ WXUEXOHQFH HQHUJ\ LV IRXU WLPHV DV ODUJH DV WKH (XOHULDQ WXUEXOHQFH HQHUJ\ DQG WKLV ZRXOG OHDG WR VHULRXV HUURU LQ XVLQJ WKHLU PRGHO 7KH RWKHU SDUDPHWHUV DUH FDOFXODWHG DV IROORZV 1HXWUDO FRQGLWLRQV &RPSOHWH KRPRJHQHLW\ LV DVVXPHG LQ ERWK WKH [ DQG WKH \ GLUHFWLRQV 7KH PHDQ YHORFLW\ SURILOH 7KH YHORFLW\ SURILOH IRU QHXWUDO FRQGLWLRQV LV RI WKH ORJDn ULWKPLF IRUP 0RQLQ DQG 2EXNKRY f 0DQ\ GHULYDWLRQV DUH DYDLOn DEOH LQ WKH OLWHUDWXUH DQG WKH UHVXOW LV LQGHSHQGHQW RI WKH PHWKRG RI GHULYDWLRQ /XPOH\ DQG 3DQRIVN\ f 7KH GHULYDWLRQ SUHVHQWHG KHUH LV EDVHG RQ GLPHQVLRQDO DQDO\VLV 7KH URXJKQHVV RI DQ DWPRVSKHULF JURXQG OD\HU LV GHVFULEHG E\ D URXJKQHVV OHQJWK ]T DOVR NQRZQ DV G\QDPLF URXJKQHVVf 7KH URXJKQHVV OHQJWK LV QRW HTXDO WR WKH DFWXDO KHLJKW RI WKH YHJHWDWLRQ
PAGE 73
RU RWKHU SK\VLFDO VWUXFWXUHV RQ WKH JURXQG EXW LV FRQVLGHUDEO\ VPDOOHU LQ PDJQLWXGH ,W YDULHV IURP WKH IUDFWLRQ RI D PP RYHU VQRZ RU VDQG WR VHYHUDO PHWHUV RYHU FLWLHV 1DZURFNL DQG 3DSD f 6LQFH WKH PL[LQJ OHQJWK LQ WKH DWPRVSKHUH LV SURSRUWLRQDO WR ]]Af WKH FKDUDFWHULVWLF OHQJWK VFDOH LV WDNHQ DV ]]Tf 7KH FKDUDFWHULVWLF YHORFLW\ LQ WKH DWPRVSKHUH LV WKH IULFWLRQ YHORFLW\ f ZKHUH LV WKH VKHDU VWUHVV DW WKH JURXQG DQG S LV WKH GHQVLW\ RI WKH DWPRVSKHUH 7KH IULFWLRQDO YHORFLW\ UHSUHVHQWV WKH GUDJ RQ WKH VXUIDFH %\ GLPHQVLRQDO DQDO\VLV WKH YHORFLW\ JUDGLHQW AXG]f LV WKHUHIRUH SURSRUWLRQDO WR X ]] f 7DNLQJ WKH FRQVWDQW RI SURSRU WLRQDOLW\ DV f§ ZH KDYH . GX G] . ]] f f ,QWHJUDWLQJ IURP ] WR ] ] ZH REWDLQ f 7KH GLPHQVLRQOHVV FRQVWDQW . LV WKH YRQ .DUUQDQ FRQVWDQW DQG LV WDNHQ 7KH URXJKQHVV OHQJWK ]T LV D SURSHUW\ RI WKH OD\HU DQG LV DV
PAGE 74
LQGHSHQGHQW RI PHWHRURORJLFDO FRQGLWLRQV 7KH IULFWLRQ YHORFLW\ X$ UHSUHVHQWV WKH PDJQLWXGH RI WKH ZLQG YHORFLW\ DQG LV LQYDULDQW ZLWK KHLJKW 7KH YDULDQFH RI WKH YHORFLWLHV 7KHVH DUH DVVXPHG WR EH HTXDO WR WKH (XOHULDQ PHDQ VTXDUH IOXFWXDWLRQV f 7KH YDULDQFH RI WKH DFFHOHUDWLRQV 7KH /DJUDQJLDQ DFFHOHUDWLRQ LV REWDLQHG IURP WKH (XOHULDQ DFFHOHUDWLRQ E\ WKH /DQJUDQJLDQ(XOHULDQ WUDQVIRUPDWLRQ XVLQJ WKH UDWLR RI WKH OHQJWK VFDOHV % .LUPVH f f 7KH VWDQGDUG GHYLDWLRQV RI WKH YHORFLW\ FDQ EH REWDLQHG HLWKHU IURP WKH H[SHULPHQWDO GDWD RU IURP FRUUHODWHG GDWD 7KH FRUUHODWLRQ RI WKH VSDWLDO GHULYDWLYH RI WKH YHORFLW\ FDQ EH REWDLQHG IURP WKH HQHUJ\ GLVVLSDWLRQ H LI RQH DVVXPHV WKDW VPDOO VFDOH VWUXFWXUH RI WKH WXUEXOHQFH LV LVRWURSLF 7KLV DVVXPSWLRQ LV YHU\ UHDVRQDEOH EHn FDXVH LVRWURS\ LV EHLQJ DVVXPHG RQO\ DW WKH GLVVLSDWLRQ VFDOH RI WXUn EXOHQFH ZKLOH WKH RYHUDOO EHKDYLRU LV VWLOO DQLVRWURSLF 7KHQ
PAGE 75
+LQ]H f e Y f 7KH HQHUJ\ GLVVLSDWLRQ KDV EHHQ IRXQG WR EH D IXQFWLRQ RI WKH ZLQG YHORFLW\ DQG KHLJKW 5HFRUG DQG &UDPHU f 5HFRUG DQG &UDPHU DGGHG WKHLU 5RXQG +LOO GDWD &UDPHU HW D7 f WR WKH FRPSRVLWH JUDSK RI ,YDQRY f 7KH GLVVLSDWLRQ UDWH DQG LWV KHLJKW GHSHQGHQFH ZHUH IRXQG WR YDU\ VLJQLILFDQWO\ ZLWK WKHUPDO VWUDWn LILFDWLRQ ,Q D ODWHU UHSRUW &UDPHU DQG 5HFRUG f ZHUH DEOH WR VXFn FHVVIXOO\ FRUUHODWH DOO WKHLU HQHUJ\ VSHFWUD PHDVXUHPHQWV E\ SORWWLQJ DOO RI WKHLU GDWD RQ D JUDSK RI GLPHQVLRQOHVV VSHFWUD YV Xr ] Q] GLPHQVLRQOHVV IUHTXHQF\ HUU 7KLV UHODWLRQVKLS JLYHV WKH KHLJKW DQG WKH YHORFLW\ GHSHQGHQFH RI WKH HQHUJ\ VSHFWUD DQG FRQVHTXHQWO\ WKH HQHUJ\ GLVVLSDWLRQ UDWH 7KH ODWWHU LV UHODWHG WR WKH HQHUJ\ VSHFWUD E\ WKH IROORZLQJ UHODWLRQ GN f R ZKHUH N LV WKH ZDYH QXPEHU LP X f
PAGE 76
DQG ( Nf e )Qf f 7KH ZDYH QXPEHU N FDQ EH QRQGLPHQVLRQDOL]HG LQ D VLPLODU PDQQHU WR WKH IUHTXHQF\ Q LI ZH OHW f 0XOWLSO\LQJ (T f E\ ] DQG VXEVWLWXWLQJ (T f LQWR LW ZH KDYH N] Q]B WU X f 6LPLODUO\ GLYLGLQJ (T f E\ XA]f DQG VLPSOLI\LQJ JLYHV LU( X ) Xr] Xr] f 6XEVWLWXWLQJ IRU ( DQG N LQ WHUPV RI ) DQG Q LQ (T f ZH REWDLQ H Y f
PAGE 77
H " WW Y Xr ] r r Q ) GQ f R WW Y Xr ] r e f ZKHUH f N 7KH GLPHQVLRQDO HQHUJ\ GLVVLSDWLRQ UDWH e LV WDNHQ DV D N N FRQVWDQW DQG LWV YDOXH REWDLQHG IURP ) YV Q FXUYHV &UDPHU DQG 5HFRUG f &RPELQLQJ (TXDWLRQV f f DQG f ZH REWDLQ DQ H[SUHVVLRQ IRU WKH VWDQGDUG GHYLDWLRQ RI WKH /DJUDQJLDQ DFFHOHUDWLRQV D %D er f D ] Y L [ 7KH PHDQ WLPH VWHSV .LUPVH f DQG *LHORZ f KDYH VKRZQ WKDW WKH PHDQ YDOXH RI WKH WLPH VWHS LV SURSRUWLRQDO WR WKH UDWLR RI WKH VWDQGDUG GHYLDWLRQV RI WKH YHORFLW\ DQG RI WKH DFFHOHUDWLRQ RU D L f
PAGE 78
ZKHUH WKH IDFWRU LV GHWHUPLQHG WKHRUHWLFDOO\ 7KH PHDQ KDJUDQJLDQ DFFHOHUDWLRQV 7KH PHDQ /DJUDQJLDQ DFFHOHUDWLRQ LQ WKH [GLUHFWLRQ DA LV WKH GHULYDWLYH RI WKH WXUEXOHQW VKHDU VWUHVV .LUPVH *LHORZ f 7KH YDULDWLRQ RI WKH VKHDU VWUHVV ZLWK KHLJKW LV H[WUHPHO\ VPDOO LQ WKH DWPRVSKHUH 0RQLQ DQG 2EXNKRY /XPOH\ DQG 3DQRIVN\ f DQG DA LV WKHUHIRUH WDNHQ DV ]HUR %HFDXVH RI QR QHW EXR\DQF\ WKH PHDQ /DQJUDQJLDQ DFFHOHUDWLRQ LQ WKH ]GLUHFWLRQ LV DOVR WDNHQ DV ]HUR 1RQQHXWUDO FRQGLWLRQV 7KH PHDQ YHORFLW\ SURILOH ,Q WKH FDVH RI QRQQHXWUDO FRQGLWLRQV RQH PXVW XVH DQ DGGLn WLRQDO OHQJWK VFDOH FDOOHG WKH 0RQLQ2EXNKRY OHQJWK / 7KH H[SUHVVLRQ IRU WKH VKHDU WKHQ FKDQJHV WR GX B 8IW G] N]]Tf > / _ ZKHUH LMf LV D XQLYHUVDO IXQFWLRQ RI / 6LQFH / IRU QHXWUDO FRQ ]] GLWLRQV f§f§ DQG WKHUHIRUH A f f
PAGE 79
7KH GLPHQVLRQOHVV VKHDU LS FDQ EH H[SDQGHG LQ D 7D\ORU VHUn LHV DERXW ]HUR DQG IRU VPDOO YDOXHV RI _ =/ _ B f RQH FDQ ZULWH L H ]] R / f 6XEVWLWXWLQJ (T f LQWR (T f DQG LQWHJUDWLQJ IURP ] WR ] ] ZH REWDLQ f 7KH 0RQLQ2EXNKRY OHQJWK LV GHILQHG DV IROORZV LQ WHUPV RI WKH WXUEXOHQW KHDW IOX[ Xr / f .67 f T& Sf . S 6LQFH WKH KHDW IOX[ LV GLIILFXOW WR REWDLQ LQ SUDFWLFH D FRQYHQLHQW ZD\ WR DYRLG WKLV GLIILFXOW\ LV E\ H[SUHVVLQJ WKH PRPHQWXP DQG KHDW IOX[HV LQ WHUPV RI WKH YHORFLW\ DQG SRWHQWLDO WHPSHUDWXUH JUDGLHQWV 7 f± GX Q A}G] f§ .K f§ 3& K G] f f
PAGE 80
ZKHUH . DQG . DUH WKH PRPHQWXP DQG KHDW WXUEXOHQW GLIIXVLYLWLHV P Q UHVSHFWLYHO\ $VVXPLQJ WKDW WKH YDULDWLRQ LQ 7 ZLWK KHLJKW LV YHU\ VPDOO DQG FDQ EH QHJOHFWHG ZH KDYH R f± GX f§ Xr . f§ S r P G] f 6XEVWLWXWLQJ (TXDWLRQV f DQG f (T f EHFRPHV GX fµrr .a Q ! / P f . P Xr GXG]` f $ JUDGLHQW VLPLODULW\ OHQJWK LV GHILQHG DV /* . P L IG? G] ef 5 f 7KHQ WKH YHORFLW\ SURILOH HTXDWLRQ FDQ EH ZULWWHQ DV B B Xr Âf R U* f IRU , 7a , /*
PAGE 81
ZKHUH LV D XQLYHUVDO FRQVWDQW 7KH FRQVWDQW ZDV WDNHQ DV IRU XQVWDEOH FRQGLWLRQV DQG IRU VWDEOH FRQGLWLRQV /XPOH\ DQG 3DQRIVN\ f )RU _=O_ ! (TXDWLRQ f LV QRW YDOLG EXW D SRZHU ODZ HTXDWLRQ &VDQDG\ f FDQ EH XVHGO f ZKHUH ]A LV DQ DUELWUDU\ KHLJKW 7KH H[SRQHQW D GHSHQGV XSRQ WKH URXJKQHVV OHQJWK ]4 WKH UHIHUHQFH KHLJKW ]A DQG WKH VWDELOLW\ 6LQFH WKH ORJOLQHDU SURILOH JLYHQ E\ (T f FDQ EH DVVXPHG WR KROG XS WR WKH KHLJKW _=/T _ WKH UHIHUHQFH KHLJKW f ZDV FKRVHQ DV WKLV YDOXH 7KHUHIRUH WKH UHIHUHQFH YHORFLW\ LV Xf± Xr OJ A ? OJ , OQ ‘ rUF . 9 R OJ f ] ! 9 f 7KH JUDGLHQW VLPLODULW\ OHQJWK GHVFULEHV WKH VWDELOLW\ ZKLOH ]T GHVFULEHV WKH SK\VLFDO FKDUDFWHULVWLFV 7KH RQO\ GLPHQVLRQOHVV SDUDPn HWHUV WKDW FDQ EH IRUPHG IURP WKHVH SDUDPHWHUV LV =r/S 7KHUHIRUH
PAGE 82
f 7KH DERYH HTXDWLRQ LV VHHQ WR EH REH\HG IRU DOO WKH 3URMHFW 3UDLULH *UDVV UXQV %DUDG f 7KH H[SRQHQW D LV SORWWHG DJDLQVW ] WKH VWDELOLW\ SDUDPHWHU /LQ )LJ f IRU DOO WKH UXQV ZLWK VWDEOH FRQGLWLRQV 7KH IXQFWLRQ D FDQ EH FRUUHODWHG IRU VWDEOH FRQn GLWLRQV E\ WKH H[SUHVVLRQ E f ] ZKHUH D LV WKH VORSH RI D YV OQ /Tf OLQH DQG E LV WKH LQWHUFHSW ] DW /J )RU XQVWDEOH FRQGLWLRQV D D FRQVWDQW f 7KH WHPSHUDWXUH SURILOH 3RWHQWLDO WHPSHUDWXUH LV GHILQHG DV WKH WHPSHUDWXUH D SDUFHO RI DLU ZRXOG DWWDLQ LI LW ZDV EURXJKW WR D VWDQGDUG UHIHUHQFH SUHVVXUH XQGHU DGLDEDWLF FRQGLWLRQV 7KH SRWHQWLDO WHPSHUDWXUH FDQ EH UHODWHG WR WKH WHPSHUDWXUH LQ WKH IROORZLQJ PDQQHU $VVXPLQJ WKH DLU WR EHKDYH DV DQ LGHDO JDV WKH ILUVW ODZ RI WKHUPRG\QDPLFV IRU DGLDEDWLF FRQGLWLRQV FDQ EH ZULWWHQ DV +HVV f 37 & G7 G3 f 3 3
PAGE 83
7 7 )LJXUH R B/ -/ -/ f¬ rr ]RWa! f¬ =4 f¯ /* 7KH YHORFLW\SURILOH H[SRQHQW D YV WKH VWDELOLW\ SDUDPHWHU ] FRQGLWLRQV n $r IRU VWDEOH
PAGE 84
,QWHJUDWLQJ (T f IURP WKH UHIHUHQFH SUHVVXUH 3 WR 3 ZH KDYH 5 S f§ 7 f &3 f ZKHUH LV WKH SRWHQWLDO WHPSHUDWXUH )RU WKH 3UDLULH *UDVV UXQV %DUDG f WKH SUHVVXUH SURILOHV ZHUH REVHUYHG WR EH OLQHDU ,Q WKLV ZRUN WKH UHIHUHQFH SUHVVXUH LV WDNHQ DV WKH SUHVVXUH DW WKH JURXQG 7KXV WKH SUHVVXUH SURILOH LV 3 3( 35 J] f ZKHUH S LV WKH GHQVLW\ DW WKH JURXQG DQG J LV WKH DFFHOHUDWLRQ . ] GXH WR JUDYLW\ LQ WKH ]GLUHFWLRQ ZKLFK LV JLYHQ E\ J] J PVHF f 6XEVWLWXWLQJ (T f LQWR (T f DQG H[SDQGLQJ LQWR D ELn QRPLDO VHULHV ZH KDYH 7^ 35 J] = 5 f 8VLQJ WKH LGHDO JDV ODZ DQG QHJOHFWLQJ WKH KLJKHU RUGHU WHUPV (T f UHGXFHV WR 7 B6B & 3 ] f
PAGE 85
7KH DERYH HTXDWLRQ LV D JRRG DSSUR[LPDWLRQ IRU (T f IRU VPDOO YDOXHV RI ] 7KH SRWHQWLDO WHPSHUDWXUH ZDV SORWWHG DJDLQVW WKH PHDQ YHORFLW\ X )LJV DQG f IRU D ODUJH QXPEHU RI 3UDLULH *UDVV VWDEOH DQG XQVWDEOH UXQV 7KH VORSHV RI WKHVH FXUYHV ZHUH REVHUYHG WR EH FRQVWDQW 7KLV LPSOLHV WKDW WKH JUDGLHQW VLPLODULW\ OHQJWK /f± FDQ EH WDNHQ DV D FRQVWDQW LQ WKH VXUIDFH OD\HU (T f 7KH 9SRWHQWLDO WHPSHUDWXUHPHDQ YHORFLW\ FXUYHV FDQ EH H[SUHVVHG DV f &RPELQLQJ (TXDWLRQV f DQG f ZH KDYH n* f $ JUDGLHQW WHPSHUDWXUH VFDOH 7 DQDORJRXV WR WKH IULFWLRQ YHORFLW\ 9M LV GHILQHG DV 7 G] ¾IF * f¯ G¾ G] f 7KH ORJOLQHDU WHPSHUDWXUH SURILOH HTXDWLRQ WKHQ LV * / ] n* f
PAGE 86
X PSV )LJXUH 7KH SRWHQWLDO WHPSHUDWXUH YV WKH PHDQ YHORFLW\ IRU VRPH 3UDLULH *UDVV ULPV ZLWK VWDEOH FRQGLWLRQV
PAGE 87
)LJXUH 7KH SRWHQWLDO WHPSHUDWXUH YV WKH PHDQ YHORFLW\ IRU VRPH 3UDLULH *UDVV UXQV ZLWK XQVWDEOH FRQGLWLRQV
PAGE 88
7KH JUDGLHQW WHPSHUDWXUH VFDOH 7 LV UHODWHG WR WKH 0RQLQ2EXNKRY 9WHPSHUDWXUH DFFRUGLQJ WR WKH IROORZLQJ HTXDWLRQ f 6LPLODUO\ WKH SRZHU ODZ WHPSHUDWXUH SURILOH LV U A f D = ! /J f 7KH YDULDQFH RI WKH YHORFLWLHV $V LQ WKH FDVH RI WKH QHXWUDO FRQGLWLRQV WKH YHORFLW\ YDULn DQFHV DUH DVVXPHG WR EH HTXDO WR WKH (XOHULDQ YDOXHV ZKLFK PD\ EH REWDLQHG IURP WKH H[SHULPHQWDO RU FRUUHODWHG GDWD 7KH YDULDQFH RI WKH DFFHOHUDWLRQV 7KH YDULDWLRQ RI WKH HQHUJ\ GLVVLSDWLRQ ZLWK VWDELOLW\ LV QRW DYDLODEOH %XW WKH IXQFWLRQDO UHODWLRQVKLS RI e ZLWK ] LV DVVXPHG Wµ EH WKH VDPH DV IRU WKH QHXWUDO FRQGLWLRQV 7KLV OHDGV WR WKH D EHLQJ LQYHUVHO\ SURSRUWLRQDO WR WKH KHLJKW 7KH FRUUHODWLRQ FR 9L HIILFLHQWV ZHUH PHDVXUHG GXULQJ WKH 3UDLULH *UDVV UXQV +DXJHQ f DQG WKHUHIRUH WKH PLFURVFDOHV RI WXUEXOHQFH FRXOG EH HVWLPDWHG )URP +LQ]H f WKH WLPH PLFURVFDOH W LV UHODWHG WR WKH FRUUHODWLRQ e
PAGE 89
RI WKH YHORFLW\ GHULYDWLYH DFFRUGLQJ WR WKH IROORZLQJ HTXDWLRQV 9 I[sf 5M Wf f L ZKHUH I[f LV WKH ORQJLWXGLQDO VSDFH FRUUHODWLRQ FRHIILFLHQW $OVR f W f $FFRUGLQJ WR 7D\ORUnV K\SRWKHVLV BB B BB W X [ f 7KHUHIRUH f 8VLQJ (TXDWLRQV f DQG f ZH REWDLQ Ya 2 9 D L f
PAGE 90
7KH PHDQ WLPH VWHSV 7KH PHDQ WLPH VWHSV DUH REWDLQHG IURP WKH WXUEXOHQW LQWHQVLW\ DQG WKH YDULDQFH RI WKH DFFHOHUDWLRQ LQ WKH VDPH PDQQHU DV IRU WKH FDVH RI QHXWUDO FRQGLWLRQV (T f 7KH PHDQ KDJUDQJLDQ DFFHOHUDWLRQV 7KH PHDQ DFFHOHUDWLRQV LQ WKH [ DQG \ GLUHFWLRQV DUH WDNHQ DV ]HUR 7KH PHDQ /DQJUDQJLDQ DFFHOHUDWLRQ LQ WKH ]GLUHFWLRQ LV REWDLQHG IURP WKH HTXDWLRQ RI PRWLRQ LQ WKH ]GLUHFWLRQ B 'Z 3 D] 'W S ] p] f /HW 3T]f GHVFULEH WKH SUHVVXUH DW VWHDG\ VWDWH QHXWUDO FRQGLWLRQV 7KHQ 3 R 35 f ,I UHSUHVHQWV WKH GHYLDWLRQ RI WKH GHQVLW\ IURP SA WKHQ f f
PAGE 91
7KH GHYLDWLRQ S FDQ EH FDOFXODWHG E\ H[SDQGLQJ S LQWR D WZRGLPHQn VLRQDO 7D\ORU VHULHV DERXW WKH SRLQW S DV IROORZV R S S S R S f ZKHUH 7 7 f $ 3 3 3 f R DQG 7A DQG 3A DUH WKH PHDQ WHPSHUDWXUH DQG SUHVVXUH SURILOHV UHn VSHFWLYHO\ IRU VWHDG\ VWDWH QHXWUDO FRQGLWLRQV $VVXPLQJ WKH LGHDO JDV ODZ DQG QHJOHFWLQJ WKH KLJKHU RUGHU WHUPV 7 e S SR S -a f R R 6XEVWLWXWLQJ (T f LQWR (T f ZH REWDLQ B S S S R R S S +R 7 + R 3 f f
PAGE 92
6XEVWLWXWLQJ (T f LQWR (T f ZH KDYH D] S G3 G] B3B 3 f 6LQFH WKH SUHVVXUH DQG WHPSHUDWXUH SURILOH DUH NQRZQ RQH FDQ WKHUHn IRUH REWDLQ DQ H[SUHVVLRQ IRU DA LQ WHUPV RI ] 7KH WHPSHUDWXUH SURn ILOH IRU WKH QHXWUDO FRQGLWLRQV LV JLYHQ E\ W H' U ] R 5 f ZKHUH ) LV WKH DGLDEDWLF ODSVH UDWH DQG LV HTXDO WR AA&Sf 6LQFH WKH UHIHUHQFH SUHVVXUH LV WDNHQ DV WKH SUHVVXUH DW WKH JURXQG OHYHO 5 75 f 6XEVWLWXWLQJ (TXDWLRQV f f f DQG f LQWR (T f ZH KDYH WKH IROORZLQJ H[SUHVVLRQ IRU 7 7 7 ,Q _ ]] ? ] f ! 7U 7 f D D f B* ] !
PAGE 93
&RPELQLQJ (TXDWLRQV f DQG f ZLWK (T f ZH KDYH DQG -/ 7 ]] ,Q * / U ] S f OQ>AM J]rr rr ? m5 K f IRU ] * 7 7 ! /* D 2' UU1L ] f f IRU ] ! B* ZKHUH r ] f r ] R f rr ] f
PAGE 94
5 f f 6LPLODUO\ WKH H[SUHVVLRQ IRU WKH SUHVVXUH GHYLDWLRQ WHUP EHFRPHV f G3 G] f $ OLQHDU DSSUR[LPDWLRQ IRU WKH DERYH HTXDWLRQ LV G] G] f rr ] 1 1[ 1 & 3 & B( 5 ] f ZKHUH f
PAGE 95
6XEVWLWXWLQJ (TXDWLRQV f f LQWR (TXDWLRQ f ZH REWDLQ r D ] $ 2 9O If 1 & /( 5 rr ] rr ] f $ 7 ZKHUH 7 LV JLYHQ E\ (TXDWLRQV f DQG f $OO WKH WHUPV LQ (T f DUH SORWWHG DJDLQVW WKH KHLJKW IRU D VHW RI W\SLFDO FRQGLWLRQV LQ )LJ f ,W FDQ EH VHHQ WKDW WKH YDULDWLRQ LQ D ] FORVH WR WKH JURXQG LV FRQVLGHUDEOH
PAGE 96
2 ] P )LJXUH 7KH YDULDWLRQ RI GLIIHUHQW WHUPV LQ WKH YHUWLFDO PHDQ /DJUDQJLDQ DFFHOHUDWLRQ HTXDWLRQ ZLWK KHLJKW 9 FR X!
PAGE 97
&+$37(5 5(68/76 &RPSXWDWLRQDO 3DUDPHWHUV 7KH 0RQWH &DUOR VLPXODWLRQ RI WKH VWRFKDVWLF PRGHO XVHG LQ WKLV ZRUN FDQ EH FDUULHG RXW IRU DQ\ DUELWUDULO\ FKRVHQ QXPEHU RI SDUWLFOHV VDPSOLQJ WLPH DQG FHOO VL]H $ VHQVLWLYLW\ DQDO\VLV ZDV FDUULHG RXW VWXG\LQJ WKH HIIHFW RI WKHVH SDUDPHWHUV DQG RSWLPDO YDOXHV WKHQ ZHUH XVHG IRU DOO WKH UXQV ,Q DGGLWLRQ RQH LV IUHH WR GLYLGH WKH GLVWULEXWLRQ IXQFWLRQV IRU WKH YHORFLWLHV DFFHOHUDWLRQV DQG WLPH VWHSV LQWR DQ\ QXPEHU RI SDUWV VHH $SSHQGL[ $f *LHORZ f WULHG GLIIHUHQW QXPEHUV RI SDUWLWLRQV LQ KLV VWXG\ DQG HYHQn WXDOO\ XVHG SDUWLWLRQV 6LQFH WKH UHTXLUHPHQWV RI SUHFLVLRQ LQ WKH DWPRVSKHUH DUH OHVV WKDQ LQ WKH ODERUDWRU\ VFDOH VLPXODWLRQV VL[WHHQ SDUWLWLRQV ZHUH DOVR XVHG LQ WKLV ZRUN IRU DOO WKH GLVWULEXn WLRQ IXQFWLRQV 7KH QXPEHU RI SDUWLFOHV ZDV YDULHG IURP WR $V RQH PD\ H[SHFW FRPSXWHU WLPH UHTXLUHPHQWV DUH URXJKO\ SURSRUWLRQDO WR WKH QXPEHU RI SDUWLFOHV UHOHDVHG 7KH VFDWWHU REVHUYHG LQ WKH UHVXOWV GHFUHDVHV ZLWK DQ LQFUHDVH LQ WKH QXPEHU RI SDUWLFOHV +RZHYHU WKH PDUJLQDO XWLOLW\ GHFUHDVHV UDSLGO\ ZLWK WKH LQFUHDVH LQ WKH QXPEHU RI SDUWLFOHV )URP WKHVH FRQVLGHUDWLRQV SDUWLFOHV ZHUH XVHG LQ DOO WKH UXQV
PAGE 98
7KH VDPSOLQJ WLPH KDV D VLPLODU HIIHFW RQ WKH UHVXOWV DV GRHV WKH QXPEHU RI SDUWLFOHV $Q LPSURYHPHQW LQ WKH UHVXOWV LV REn VHUYHG LI WKH SDUWLFOHV DUH VDPSOHG PRUH RIWHQ LH LI WKH VDPSOLQJ WLPH LV GHFUHDVHG )URP D VLPLODU DQDO\VLV D VDPSOLQJ WLPH RI VHFRQGV ZDV VHOHFWHG 7KH DYDLODEOH FRPSXWHU VWRUDJH LV DQ REYLRXV OLPLWDWLRQ RQ WKH WRWDO QXPEHU RI FHOOV ,I WKH GRZQZLQG GLVWDQFH WKDW WKH SDUWLn FOHV KDYH WR WUDYHO LQFUHDVHV WKHQ WKH FHOO VL]H PXVW EH LQFUHDVHG $OVR VLQFH WKH VSUHDG LQFUHDVHV DSSUR[LPDWHO\ OLQHDUO\ ZLWK WKH GRZQn ZLQG GLVWDQFH WKH FHOO VL]H ZDV YDULHG ZLWK WKH GRZQZLQG GLVWDQFH DQG LQVWHDG RI D FRPSOHWH QHWZRUN RI FHOOV WZRGLPHQVLRQDO FHOO VWUXFWXUHV ZHUH XVHG DW GLIIHUHQW GRZQZLQG SRVLWLRQV 6LQFH SHUIHFW V\PPHWU\ LV DVVXPHG LQ WKH \GLUHFWLRQ DOO WKH SDUWLFOH SRVLWLRQV DUH IROGHG LQ WKH \GLUHFWLRQ 7KH D[LV FRUUHVSRQGV WR WKH FHQWHU RI WKH ILUVW FHOO DQG LV WKHUHIRUH RQH KDOI WKH VL]H RI WKH RWKHU FHOOV 7KH FHOO VL]H KDV QR HIIHFW RQ WKH UHVXOWV H[FHSW ZKHQ DQ H[n WUHPHO\ ODUJH FHOO VL]H LV XVHG FORVH WR WKH VRXUFH 9DOLGDWLRQ RI WKH 0RGHO 7KH PRGHO ZDV YDOLGDWHG E\ FRPSDULQJ WKH UHVXOWV RI D VLPXn ODWLRQ ZLWK WKRVH REWDLQHG LQ WKH 3UDLULH *UDVV GLIIXVLRQ H[SHULPHQWV %DUDG +DXJHQ f 7KH FRUUHODWHG WXUEXOHQFH GDWD RI &UDPHU DQG 5HFRUG f ZHUH XVHG WR VWXG\ WKH HIIHFWV RI GLIIHUn HQW SK\VLFDO DQG PHWHRURORJLFDO SDUDPHWHUV LQFOXGLQJ WKH HIIHFWV RI QHJOHFWLQJ WKH D[LDO GLIIXVLRQ 7KH UHVXOWV DUH FRUUHODWHG LQ D XVHn IXO IRUP DQG JUDSKV IRU FRUUHODWHG GLVSHUVLRQ FRHIILFLHQWV DQG HGG\
PAGE 99
GLIIXVLYLWLHV DUH SORWWHG 1HXWUDO FRQGLWLRQV 3UDLULH *UDVV GLIIXVLRQ H[SHULPHQWV %DUDG +DXn JHQ f ZHUH FRQGXFWHG DW 2n1HLOO 1HEUDVND 6XOIXU GLR[LGH ZDV UHOHDVHG DW D KHLJKW RI P IURP WKH JURXQG DQG FRQFHQWUDWLRQV ZHUH PHDVXUHG LQ WKH KRUL]RQWDO GLUHFWLRQ DORQJ ILYH FRQFHQWULF DUFV DQG LQ WKH YHUWLFDO GLUHFWLRQ RQ VL[ WRZHUV 7KH ILYH DUFV ZHUH VHPLn FLUFXODU DQG RI UDGLL DQG PHWHUV $OO WKH VL[ WRZHUV ZHUH DORQJ WKH P DUF )RU WKH VDPH FRQGLWLRQV DV WKRVH RI 5XQ ZKLFK ZDV DSSUR[LPDWHO\ QHXWUDOf WKH FRQFHQWUDWLRQ GLVWULn EXWLRQV LQ WKH \GLUHFWLRQ DW WKH P P DQG P DUFV DQG LQ WKH ]GLUHFWLRQ IRU WZR WRZHUV \ P DQG \ Pf DUH FRPSDUHG LQ )LJXUHV f WR f UHVSHFWLYHO\ ZLWK WKH H[SHULPHQWDO UHn VXOWV DQG ZLWK WKH VLPXODWLRQ UHVXOWV RI %XOOLQ DQG 'XNOHU f 7KH FRQFHQWUDWLRQ SURILOHV REWDLQHG IURP WKH 3DVTXLOO*LIIRUG HTXDn WLRQ 7XUQHU f DUH DOVR SORWWHG RQ WKH VDPH JUDSK XVLQJ WKH VWDQGDUG GHYLDWLRQV IRU WKH KRUL]RQWDO DQG YHUWLFDO GLVSHUVLRQ IURP WKH 3DVTXLOO*LIIRUG FKDUWV FODVV 'f ,Q DOO WKH FDVHV FRPSOHWH KRPRJHQHLW\ LQ ERWK WKH [ DQG WKH \ GLUHFWLRQV LV DVVXPHG ,W FDQ EH VHHQ WKDW WKH UHVXOWV DUH VLJQLILFDQWO\ EHWWHU WKDQ XVLQJ HLWKHU RI WKH WZR DERYH PHQWLRQHG PRGHOV 5XQ QXPEHU ZDV VHHQ WR EH WKH FORVHVW WR SHUIHFWO\ QHXWUDO FRQGLWLRQV 7KH UHVXOWV DW GLIIHUHQW GRZQZLQG GLVWDQFHV LQ WKH \ DQG ]GLUHFWLRQV DUH SORWWHG LQ )LJV f WR f 7KH PHDVXUHG YDOXHV RI WKH YHORFLW\ SURILOH DQG WKH WXUEXOHQW LQWHQVLWLHV ZHUH XVHG LQ
PAGE 100
PJ P FR 7 7 1HDU QHXWUDO FnRQGLWLRQV /* Pf [ LQ ] P ? 9 k ([SHULPHQWDO 2 7KLV ZRUN %XOOLQ t 'XNOHU f B 3DVTXLOO*LIIRUG FODVV 'f ? !? k [ r 2 9 ? ? Rn 4 R rn R A 9 ? n R ; R k , B/ \ P &RQFHQWUDWLRQ GLVWULEXWLRQ LQ WKH \GLUHFWLRQ IRU 3URMHFW 3UDLULH *UDVV 5XQ DW [ )LJXUH
PAGE 101
a 1HDU QHXWUDO FRQGLWLRQV /J Pf [ P ] P 2 H ? k ([SHULPHQWDO 27KLV ZRUN f§ %XOOLQ t 'XNOHU f 3DVTXLOO*LIIRUG FODVV 'f 2 [ k ? n? ? ? r 2 2 ? ? ? ? R ? ?R ? ? k ? k f± k R f§ RB -6Âf§kS \ P )LJXUH &RQFHQWUDWLRQ GLVWULEXWLRQ LQ WKH \GLUHFWLRQ IRU 3URMHFW 3UDLULH *UDVV 5XQ DW [ P
PAGE 102
E2 &f§ 1HDU QHXWUDO FRQGLWLRQV 2 U /J Pf k ([SHULPHQWDO [ P 27KLV ZRUN ] P f§%XOOLQ t 'XKOHU f f§ R R f§ 3DVTXLOO*LIIRUG FODVV 'f ? 2k ? 2 f§ k A ? 7 k ?Q k n ? ? k ? ? } ? ? ? ? ? ? ? 6 ? ? ? 9 ? r G f§ 9 ? 9 \ ? \ ? ! \ k n 6 L / / 4 m2 \ P f± )LJXUH &RQFHQWUDWLRQ GLVWULEXWLRQ LQ WKH \GLUHFWLRQ IRU 3UDLULH *UDVV 5XQ DW [ P
PAGE 103
UR *2 H 7 7 ?AR ?? k? ? n R ? ? X V 1HDU QHXWUDO FRQGLWLRQV /J Pf [ P \ P k ([SHULPHQWDO 2 7KLV ZRUN %XOOLQ t 'XKOHU f 3DVTXLOO*LIIRUG FODVV 'f ? ? k 1 Vr ? ? ? ? ? ? ? fµ? ? 61r ? 6V2 ? n* ] P )LJXUH &RQFHQWUDWLRQ GLVWULEXWLRQ LQ WKH ]GLUHFWLRQ IRU 3UDLULH *UDVV 5XQ DW [ P 92
PAGE 104
c UR WR D 1HDU QHXWUDO FRQGLWLRQV /* Pf [ P ] P 7 2 k n 3 ? 2 ([SHULPHQWDO 2 7KLV ZRUN f§3DVTXLOO *LIIRUG FODVV 'f O ? ? ? ? ? ? k n2 R? 21 k 2 ? R ? k ? k Z R f§f§O R L \ P )LJXUH &RQFHQWUDWLRQ GLVWULEXWLRQ LQ WKH \GLUHFWLRQ IRU 3URMHFW 3UDLULH *UDVV 5XQ DW [ UD 1
PAGE 105
7 7 7 &2 1HDU QHXWUDO FRQGLWLRQV /J Pf [ P ] P r k -f§2f§kkm t R R p ([SHULPHQWDO 2 7KLV ZRUN f§ 3DVTXLOO*LIIRUG FODVV 'f k k n ? \ ?F ? ? YAk ? R b ? R R , \ P )LJXUH &RQFHQWUDWLRQ GLVWULEXWLRQ LQ WKH \GLUHFWLRQ IRU 3URMHFW 3UDLULH *UDVV 5XQ DW [ P
PAGE 106
\ P )LJXUH &RQFHQWUDWLRQ GLVWULEXWLRQ LQ WKH \GLUHFWLRQ IRU 3URMHFW 3UDLULH *UDVV 5XQ DW [ P
PAGE 107
)LJXUH &RQFHQWUDWLRQ GLVWULEXWLRQ LQ WKH ]GLUHFWLRQ IRU 3URMHFW 3UDLULH *UDVV 5XQ DW [ P
PAGE 108
WKH VLPXODWLRQV 7KH VFDWWHU LQ WKH REVHUYHG YDOXHV LQGLFDWH WKH PDJn QLWXGH RI HUURU LQYROYHG LQ WKH PHDVXUHPHQWV 7KH VLPXODWLRQ UHVXOWV DUH ZLWKLQ WKH HUURU LQYROYHG 7KH 3DVTXLOO*LIIRUG FKDUWV GR QRW WDNH LQWR DFFRXQW WKH YDULDWLRQV LQ WKH WHUUDLQ VRXUFH KHLJKW DQG WKH WXUn EXOHQW LQWHQVLWLHV 7KH\ UHSUHVHQW WKH DYHUDJH FRQGLWLRQV H[LVWLQJ LQ WKH DWPRVSKHUH 7KH GLIIXVLRQ PRGHO SUHVHQWHG LQ WKLV ZRUN H[SOLn FLWO\ WDNHV FDUH RI WKH YDULDWLRQV LQ WKH SDUDPHWHUV LQYROYHG DQG LV WKXV D FRQVLGHUDEOH LPSURYHPHQW RYHU WKH XVH RI WKH 3DVTXLOO*LIIRUG FKDUWV 1RQQHXWUDO FRQGLWLRQV ,W LV VHHQ WKDW VHYHUDO RI WKH SDUDPHWHUV WKDW FRQVWLWXWH DQ LQSXW IRU WKH VLPXODWLRQ DUH IXQFWLRQV RI WKH VWDELOLW\ 7KHVH SDUDPHWHUV LQFOXGH WKH WXUEXOHQFH LQWHQVLWLHV WKH VWDQGDUG GHYLDWLRQ RI WKH DFFHOHUDWLRQV WKH PHDQ YDOXH RI WKH WLPH VWHSV WKH PHDQ YHORFn LW\ SURILOH DQG WKH PHDQ /DJUDQJLDQ DFFHOHUDWLRQV )URP WKH DYDLODEOH 3UDLULH *UDVV PHDVXUHPHQWV LW LV SRVVLEOH WR HYDOXDWH DOO WKH SDUDPHn WHUV ZLWK WKH H[FHSWLRQ RI WKH PHDQ /DJUDQJLDQ DFFHOHUDWLRQ LQ WKH YHUWLFDO GLUHFWLRQ D ] 7KH QHW IRUFH DFWLQJ RQ D SDUWLFOH LQ WKH YHUWLFDO GLUHFWLRQ LV WKH GLIIHUHQFH LQ WKH SUHVVXUH JUDGLHQW IRUFH DQG WKH JUDYLDWLRQDO IRUFH 7KLV GLIIHUHQFH LV YHU\ VPDOO DQG LV FRQVHTXHQWO\ YHU\ GLIILn FXOW WR HVWLPDWH +RZHYHU UHDVRQDEOH DJUHHPHQW FDQ VWLOO EH REWDLQHG HYHQ LI WKH PHDQ /DJUDQJLDQ DFFHOHUDWLRQV DUH WDNHQ DV ]HUR 7R LOOXVWUDWH WKH SUHFLVLRQ LQ WKH HYDOXDWLRQ RI D FRQVLGHU ] WKH SUHVVXUH JUDGLHQW PHDVXUHPHQWV IRU 5XQ QXPEHUV DQG RI
PAGE 109
3URMHFW 3UDLULH *UDVV 7ZR LQGHSHQGHQW PHDVXUHPHQWV RI WKH SUHVVXUH JUDGLHQW GXULQJ WKH 3UDLULH *UDVV H[SHULPHQWV %DUDG f ZHUH PDGH XVLQJ WKH 5DZLQVRQGH DQG DLUFUDIW VDPSOLQJ WHFKQLTXHV 7KHVH JLYH WKH IROORZLQJ DFFHOHUDWLRQV DW WKH VWDFN KHLJKW H[SUHVVHG DV SHU FHQW RI J 5XQ 1R 5DZLQVRQGH $LUFUDIW $V FDQ EH VHHQ LW LV QRW SRVVLEOH WR XVH HLWKHU RI WKH WZR SUHVn VXUH JUDGLHQW PHDVXUHPHQWV WR FDOFXODWH WKH PHDQ DFFHOHUDWLRQ 7R VWXG\ WKH HIIHFW RI WKH YHUWLFDO /DJUDQJLDQ DFFHOHUDWLRQ RQ WKH FRQFHQWUDWLRQ GLVWULEXWLRQ WKH DFFHOHUDWLRQ ZDV WDNHQ DV ]HUR RQH SHU FHQW RI J DFFHOHUDWLRQ GXH WR JUDYLW\f DQG WZR SHU FHQW RI J 5XQ QXPEHUV / Pf DQG / Pf FDQ EH FRQ VLGHUHG DV EHLQJ UHSUHVHQWDWLYH RI WKH XQVWDEOH DQG VWDEOH FRQGLWLRQV UHVSHFWLYHO\ 7KH UHVXOWV IRU 5XQ DUH VKRZQ LQ )LJXUHV f WR f DQG IRU 5XQ LQ )LJXUHV f WR f 7KHVH VLPXODWLRQV ZHUH FDUULHG RXW XVLQJ WKH PHDQ WLPH VWHSV REWDLQHG IURP WKH FRUUHODn WLRQ RI &UDPHU DQG 5HFRUG f RI HQHUJ\ GLVVLSDWLRQ IRU QHXWUDO FRQn GLWLRQV $OWKRXJK WKH FRQFHQWUDWLRQ GLVWULEXWLRQV DUH VHHQ WR FKDQJH ZKHQ WKH PHDQ /DJUDQJLDQ DFFHOHUDWLRQ LQ WKH YHUWLFDO GLUHFWLRQ DA LV YDULHG IURP ]HUR WR J DQG J WKH DEVROXWH PDJQLWXGH RI D LV H[SHFWHG WR EH YHU\ VPDOO D VPDOO IUDFWLRQ RI Jf ]
PAGE 110
)LJXUH 7KH PHDQ /DJUDQJLDQ DFFHOHUDWLRQ HIIHFW IRU 3URMHFW 3UDLULH *UDVV 5XQ DW [ P
PAGE 111
)LJXUH 7KH PHDQ /DJUDQJLDQ DFFHOHUDWLRQ HIIHFW IRU 3URMHFW 3UDLULH *UDVV 5XQ DW [ P
PAGE 112
)LJXUH 7KH PHDQ /DJUDQJLDQ DFFHOHUDWLRQ HIIHFW IRU 3URMHFW 3UDLULH *UDVV 5XQ LW DW [ P
PAGE 114
7KH PHDQ /DJUDQJLDQ DFFHOHUDWLRQ HIIHFW IRU 3URMHFW 3UDLULH *UDVV 5XQ DW [ P )LJXUH
PAGE 115
< P )LJXUH 7KH PHDQ /DJUDQJLDQ DFFHOHUDWLRQ HIIHFW IRU 3URMHFW 3UDLULH *UDVV 5XQ DW [ P
PAGE 116
] P 7KH PHDQ /DQJUDQJLDQ DFFHOHUDWLRQ VHQVLWLYLW\ IRU 3URMHFW 3UDLULH *UDVV 5XQ DW [ P )LJXUH
PAGE 117
7KXV WKH PHDQ /DJUDQJLDQ DFFHOHUDWLRQ PD\ EH WDNHQ DV ]HUR ZLWKRXW LQWURGXFLQJ DQ\ VHULRXV HUURUV LQ WKH UHVXOWV 7KH VLPXODWLRQ SDUDPHWHUV ZHUH FDOFXODWHG DV VKRZQ LQ 6HFWLRQ f IRU QRQQHXWUDO FRQGLWLRQV 7KH YDULDQFHV RI WKH /DJUDQJLDQ DFFHOHUDWLRQV ZHUH REWDLQHG IURP WKH PLFURVFDOHV RI WXUEXOHQFH W 6HH ( (T f 7KH PLFURVFDOHV ZHUH HYDOXDWHG E\ ILWWLQJ RVFXODWLQJ SDUDn ERODV WR WKH DXWRFRUUHODWLRQ FXUYHV FORVH WR W 6HH )LJ f 6LQFH WKH FRUUHODWLRQ FRHIILFLHQWV ZHUH PHDVXUHG DW D KHLJKW RI P RQO\ WKH PLFURVFDOHV FRXOG EH FDOFXODWHG DW WKLV KHLJKW RQO\ 7KH YDULDQFHV RI WKH /DJUDQJLDQ DFFHOHUDWLRQV DQG WKH PHDQ WLPH VWHSV ZHUH DVVXPHG WR YDU\ ZLWK KHLJKW LQ WKH VDPH ZD\ DV IRU WKH QHXWUDO N FRQGLWLRQV $VVXPLQJ R DQG e WR EH FRQVWDQW (T f FDQ EH YL ZULWWHQ DV FRQVW f§ D ] L f DQG IURP (T f ZH KDYH FRQVW ] 7KHUHIRUH DQG f f f
PAGE 118
5HWf WKH \GLUHFWLRQ FORVH WR [ IRU 5XQ
PAGE 119
0RQWH &DUOR VLPXODWLRQV ZHUH FDUULHG RXW IRU 5XQ QXPEHUV DQG DQG WKH UHVXOWV ZKLFK IRU GLIIHUHQW GRZQZLQG GLVWDQFHV DUH VKRZQ LQ )LJXUHV f WR f DUH IRXQG WR EH VDWLVIDFWRU\ $OVR LW VKRXOG EH QRWHG WKDW DJUHHPHQW ZLWK WKH H[SHULPHQWDO GDWD FRXOG EH LPSURYHG LI WKH YHUWLFDO /DJUDQJLDQ DFFHOHUDWLRQ FRXOG EH FDOFXODWHG ZLWK EHWWHU SUHFLVLRQ &RUUHODWHG 5HVXOWV 7KH FRUUHODWHG GLPHQVLRQOHVV WXUEXOHQW LQWHQVLW\ GDWD D X f RI &UDPHU DQG 5HFRUG f IRU QHXWUDO FRQGLWLRQV ZHUH YL XVHG WR VWXG\ WKH HIIHFW RI GLIIHUHQW SDUDPHWHUV DQG WR FRUUHODWH WKH UHVXOWV IRU SUDFWLFDO XVH 7KH IULFWLRQ YHORFLW\ X$ DSSHDUV DV D FRQVWDQW YHORFLW\ LQ WKH YHORFLW\ SURILOH HTXDWLRQV DQG LV D PHDVn XUH RI WKH VKHDU VWUHVV DW WKH JURXQG $OWKRXJK WKH WXUEXOHQW LQWHQn VLWLHV FKDQJH ZKHQ XA FKDQJHV WKH DYHUDJH YHORFLW\ FKDQJHV E\ WKH VDPH IDFWRU DQG WKHUHIRUH WKH GLVSHUVLRQ FRHIILFLHQWV VWDQGDUG GHYLn DWLRQV RI WKH FRQFHQWUDWLRQ GLVWULEXWLRQVf UHPDLQ XQFKDQJHG 7KHUHn IRUH WKH IULFWLRQ YHORFLW\ ZDV XVHG DV D FKDUDFWHULVWLF YHORFLW\ IRU WKH JURXQG OD\HU LQ WKH DWPRVSKHUH 7KH WXUEXOHQW LQWHQVLW\ LQ WKH D[LDO PHDQ ZLQGf GLUHFWLRQ LV KLJKHU WKDQ LQ WKH RWKHU WZR GLUHFWLRQV EXW WKH GLVSHUVLRQ FRHIILn FLHQWV UHPDLQ SUDFWLFDOO\ XQFKDQJHG ZKHQ WKH D[LDO GLIIXVLRQ LV QHJOHFWHG 6HH )LJ f 7KH 3DVTXLOO*LIIRUG HTXDWLRQ GRHV QRW FRQWDLQ WKH D[LDO GLIIXVLRQ WHUP $ PRGLILHG IRUP RI WKH 3DVTXLOO *LIIRUG HTXDWLRQ LV SURSRVHG LQ WKLV ZRUN DQG WKH D[LDO GLIIXVLRQ
PAGE 120
k kR &2 V 4 k 2 8QVWDEOH FRQGLWLRQV /J Pf [ P ] P 2 ([SHULPHQWDO 2 7KLV ZRUN 2 2 k k k k F3 i k k k k , 4 k mR P )LJXUH &RQFHQWUDWLRQ GLVWULEXWLRQ LQ WKH \GLUHFWLRQ IRU 3URMHFW 3UDLULH *UDVV 5XQ DW [ P
PAGE 121
&2 f k B 8QVWDEOH FRQGLWLRQV /J Pf [ P ] P * ([SHULPHQWDO 2 7KLV ZRUN k t 2 H F k k k H k k k k R k ; ; k k k k -; H3 _<_ P )LJXUH &RQFHQWUDWLRQ GLVWULEXWLRQ LQ WKH \GLUHFWLRQ IRU 3URMHFW 3UDLULH *UDVV 5XQ DW [ P
PAGE 122
2 f¯ k k E k 8QVWDEOH FRQGLWLRQV /J Pf [ P ] P k ([SHULPHQWDO 2 7KLV ZRUN &2 RR k k k k k 2 k k k fü! k k k k k k bs k kL2 \ } P )LJXUH &RQFHQWUDWLRQ GLVWULEXWLRQ LQ WKH \GLUHFWLRQ IRU 3URMHFW 3UDLULH *UDVV 5XQ DW [ P
PAGE 123
PJPn 8QVWDEOH FRQGLWLRQV /J Pf [ P \ P k ([SHULPHQWDO 2 7KLV ZRUN a k Rk R f§ k U f§ 2 f§ k R 2 f§ k r 2 B r &12 ] P )LJXUH &RQFHQWUDWLRQ GLVWULEXWLRQ LQ WKH ]GLUHFWLRQ IRU 3URMHFW 3UDLULH *UDVV 5XQ DW [ P ,OO
PAGE 124
0 D R , _ p R p n R 8QVWDEOH FRQGLWLRQV /* Pf [ P k ([SHULPHQWDO \ P 2 7KLV ZRUN f§ R R k f§ R k 2 2 k f§ R r R B] P )LJXUH &RQFHQWUDWLRQ GLVWULEXWLRQ LQ WKH ]GLUHFWLRQ IRU 3URMHFW 3UDLULH *UDVV 5XQ DW [ P
PAGE 125
L aUaf§L f§ c If§ f§ U 6WDEOH FRQGLWLRQV k ([SHULPHQWDO /* Pf k [ P r k 2 7KLV ZRUN ] P 2 k 2 R k R 2 !4f§f§t4 kR R \ P k 2 &2 ‘/kf§eecf§T XUH &RQFHQWUDWLRQ GLVWULEXWLRQ LQ WKH \GLUHFWLRQ IRU 3URMHFW 3UDLULH *UDVV 5XQ DW [ P
PAGE 126
af§U f§ L U f§c _ > c L 6WDEOH FRQGLWLRQV / Pf k ([SHULPHQWDO * k k [ P ] P 2 7KLV ZRUN f§ f§ R R R k k 2 2 f§ f§ R k R fµ k f§ R R k R k QrO 2 p_r _ _ , L R k r IIO2 % E H \ P )LJXUH &RQFHQWUDWLRQ GLVWULEXWLRQ LQ WKH \GLUHFWLRQ IRU 3URMHFW 3UDLULH *UDVV 5XQ DW [ P
PAGE 127
R aL Ua aUa f§ 6WDEOH FRQGLWLRQV /* Pf R 2 ([SHULPHQWDO [ P k k BB ] P R R 2 7KLV ZRUN 2 R k k 2 R 2 2 B R R R L R O R K \ P &RQFHQWUDWLRQ GLUHFWLRQ LQ WKH \GLUHFWLRQ IRU 3URMHFW 3UDLULH *UDVV 5XQ DW [ P )LJXUH
PAGE 128
7 7 7 k 6WDEOH &RQGLWLRQV /J Pf k ([SHULPHQWDO [ P \ P 2 7KLV ZRUN 2 k R R R R B R k ? R / / L ] P &RQFHQWUDWLRQ GLVWULEXWLRQ LQ WKH ]GLUHFWLRQ IRU 3URMHFW 3UDLULH *UDVV 5XQ DW [ P
PAGE 129
2n ]r R 2 LQFOXGLQJ D[LDO GLIIXVLRQ 2 QHJOHFWLQJ D[LDO GLIIXVLRQ \ b R R R R rr Rrr R f§ R r R f§ BO [ P )LJXUH (IIHFW RI QHJOHFWLQJ WKH D[LDO GLIIXVLRQ RQ WKH GLVSHUVLRQ FRHIILFLHQW LQ WKH \GLUHFWLRQ
PAGE 130
WHUP LV RPLWWHG +RZHYHU DOO WKH VLPXODWLRQV ZHUH FDUULHG RXW LQn FOXGLQJ WKH UDQGRP IOXFWXDWLRQV LQ WKH D[LDO GLUHFWLRQ DQG WKH UHn VXOWLQJ GLIIXVLYLWLHV DQG GLVSHUVLRQ FRHIILFLHQWV FDQ EH FRQVLGHUHG DV HIIHFWLYH YDOXHV LH YDOXHV ZKLFK WDNH LQWR DFFRXQW ZKDWHYHU HIIHFW D[LDO GLIIXVLRQ KDV RQ WKH FURVV ZLQG DQG WKH YHUWLFDO GLIIX VLYLWLHV DQG RQ WKH FRUUHVSRQGLQJ GLVSHUVLRQ FRHIILFLHQWV 7KH 3DVTXLOO*LIIRUG FKDUWV IDLO WR WDNH LQWR DFFRXQW WKH YDULDWLRQV LQ WKH VWDFN KHLJKW + DQG WKH URXJKQHVV OHQJWK ]T 6LQFH WKH DYHUDJH HGG\ VL]H LQ WKH DWPRVSKHUH LV SURSRUWLRQDO WR WKH GLVn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f DQG f 6LQFH WKH UHVXOWV DUH VHHQ WR EH QHDUO\ *DXVVLDQ LQ WKH \GLUHF WLRQ DQG QR ODUJH GHYLDWLRQV IURP *DXVVLDQ GLVWULEXWLRQ DUH VHHQ LQ WKH ]GLUHFWLRQ WKH IROORZLQJ PRGLILHG 3DVTXLOO*LIIRUG HTXDWLRQ ZDV XVHG WR FRUUHODWH WKH VLPXODWLRQ UHVXOWV
PAGE 131
;P )LJXUH (IIHFW RI VWDFN KHLJKW RQ WKH VWDQGDUG GHYLDWLRQ RI WKH FRQFHQWUDWLRQ GLVWULEXWLRQ LQ WKH \GLUHFWLRQ
PAGE 132
)LJXUH (IIHFW RI URXJKQHVV RQ WKH VWDQGDUG GHYLDWLRQ RI WKH FRQn FHQWUDWLRQ GLVWULEXWLRQ LQ WKH \GLUHFWLRQ
PAGE 133
F H[S ZKHUH F R LUD R \ ] X f DQG 8M LV WKH YHORFLW\ QHHGHG WR PDLQWDLQ D PDVV EDODQFH DW HDFK [ 6LQFH WKH DWPRVSKHUH LV XQERXQGHG LW LV GLIILFXOW WR GHILQH D ILQLWH RYHUDOO DYHUDJH YHORFLW\ +RZHYHU RQH FDQ GHILQH WKH FRQYHFWLRQ YHORFLW\ IRU WKH SOXPH RQ WKH EDVLV RI WKH PDVV EDODQFH DV 8)[f &2 &2 FX G\ G] 22 &2 F G\ G] f (T f LV D FXUYH ILW WR WKH DFWXDO GLVWULEXWLRQV DQG D PRUH ULJRURXV DQDO\VLV VKRXOG WDNH LQWR DFFRXQW WKH GHYLDWLRQV IURP WKH *DXVVLDQ EHKDYLRU LQ WKH YHUWLFDO GLUHFWLRQ 7KLV LV GLVn FXVVHG LQ 6HFWLRQ f $Q LQWHJUDO PDVV EDODQFH DFFRUGLQJ WR (T f ZDV FDUULHG RXW IRU DOO WKH UXQV DQG WKH FRQYHFWLRQ YHORFLW\ 8 ZDV HYDOXDWHG U DV D IXQFWLRQ RI WKH GRZQZLQG GLVWDQFH [ 7KH VWDQGDUG GHYLDWLRQV
PAGE 134
RI WKH FRQFHQWUDWLRQ GLVWULEXWLRQ LQ WKH \ DQG WKH ]GLUHFWLRQV ZHUH FDOFXODWHG E\ LQWHJUDWLQJ WKH FRQFHQWUDWLRQ GLVWULEXWLRQ HTXDWLRQ f LQ WKH ] DQG WKH \GLUHFWLRQ UHVSHFWLYHO\ 7KHQ , \ 4 W)a D 8 ] ) H[S f ZKHUH f DQG DOVR , ] W7 D 8B \ ) f ZKHUH &2 f 7KH FRQFHQWUDWLRQ LQWHJUDOV , DQG , ZHUH FDOFXODWHG IURP \ ] (TXDWLRQV f DQG f UHVSHFWLYHO\ 7KXV (TXDWLRQV f DQG f KDYH RQO\ RQH XQNQRZQ SDUDPHWHU HDFK YL] ¾DQG 2A $ OHDVW VTXDUH FXUYH ILW PHWKRG ZDV XVHG WR HYDOXDWH 2 DQG 2 7KH YDOXHV RI \
PAGE 135
DQG ¾] DUH WKHUHIRUH WKH SDUDPHWHUV ,Q WKH OHDVW VTXDUH FXUYH ILW RI WKH *DXVVLDQ HTXDWLRQ -XVW DV HGGLHV LQ D SLSH KDYH PDFURVFDOHV GHWHUPLQHG E\ WKH GLDPHWHU VR DOVR WKH VRXUFH KHLJKW LV LQGLFDWLYH RI WKH PD[LPXP YHUn WLFDO HGG\ VL]H LQ WKH DWPRVSKHUH DQG LV WKHUHIRUH WDNHQ DV WKH FKDUn DFWHULVWLF OHQJWK VFDOH 8VLQJ WKH FKDUDFWHULVWLF YHORFLW\ X$ DQG WKH FKDUDFWHULVWLF OHQJWK + WKH HTXDWLRQ IRU WKH FRQFHQWUDWLRQ GLVn WULEXWLRQ LV QRQGLPHQVLRQDOL]HG ]rOf_ . f ZKHUH r F FX + r 4 f¯ f r F R F 8r+ R r 4 O WW¾ rR r8B \ ] ) r f f r \ f r ] ]B + f
PAGE 136
f f f 7KH YHORFLW\ SURILOH IRU QHXWUDO FRQGLWLRQV EHFRPHV r X X ,Q ] rr f ZKHUH ]R R + f ,W LV VHHQ WKDW GLIIHUHQW ]A YDOXHV JLYH GLIIHUHQW FXUYHV MOU A A ZKHQ 2 YHUVXV [ LV SORWWHG IRU HDFK RI WKH ] YDOXHV 7KHVH \ R FXUYHV IRU HDFK YDOXH RI ] DUH VKRZQ LQ )LJXUHV f WR f R ,W FDQ EH VHHQ WKDW LQGLYLGXDO YDULDWLRQV LQ ] DQG + DUH RI QR FRQ R VHTXHQFH DV ORQJ DV WKHLU UDWLR UHPDLQV FRQVWDQW )RU D VPDOO UDQJH N N N RI [ 2 DQG D VLPXODWLRQ UHVXOWV DUH SORWWHG LQ )LJXUHV f DQG \ ] N f IRU DOO WKH ] YDOXHV VWXGLHG 7KH HIIHFW RI WKH SDUDPHWHU R ] LV ZHOO LOOXVWUDWHG LQ D FURVV SORW RI YHUVXV ]n IRU WZR YDOXHV R \ R N RI WKH GLPHQVLRQOHVV GRZQ ZLQG GLVWDQFH [ )LJ f 7KH URXJKQHVV HIIHFW LV TXLWH VLJQLILFDQW DQG FDQ QRW EH QHJOHFWHG LQ WKH DWPRVSKHULF
PAGE 137
Q U f§QUU f§U 7a,f§ aL cf§U ]r R f§ ] Rf¬ +P XrPSV R f§ R R f§ R fµ R 2 R R fµ f§ 2 f§ a fµ O , f¯ [ )LJXUH 7KH KRUL]RQWDO GLVSHUVLRQ FRHIILFLHQW IRU ]T
PAGE 138
r ' aU Qf§ f¯Lf§P f§ L P f§ f§Lf§Ua r ] R ] P R +P X$PSV R R R R R f§ 2 f§ f§ 2 2 fµ / , L O U O , L L LR r )LJXUH 7KH KRUL]RQWDO GLVSHUVLRQ FRHIILFLHQW IRU ] r R
PAGE 139
L cf§P ]r R ]4P +P XrPSV L f§P R R R 2 L B/, , ,,, r [ )LJXUH 7KH KRUL]RQWDO GLVSHUVLRQ FRHIILFLHQW IRU ]r
PAGE 140
]r 2222, =P + P 8UF PSV R Y Â’ R B/B B/ Â’ Â’ t k )LJXUH 7KH KRUL]RQWDO GLVSHUVLRQ FRHIILFLHQW IRU ]
PAGE 141
. ! 2 Q U UU L U LU 7_ f§M f§UQU r ] R B ] R +WQ 8r}P36 R R R R 2 R 2 f§ 2 / , , / /B/B r [ )LJXUH 7KH KRUL]RQWDO GLVSHUVLRQ FRHIILFLHQW IRU ]r
PAGE 142
L77 L U r =P +P 8APSV R R 9 Y k k , / k )LJXUH 7KH KRUL]RQWDO GLVSHUVLRQ FRHIILFLHQW IRU ]
PAGE 143
a -a 8 ] + ] $ R R 2 ’ k + 2 2 ! 9 $ D k R µ R 2 A 2 k ( $ ( N 9 $ R 9 t $ R 9 [ )LJXUH $ FRPSDULVRQ RI D YV [ IRU GLIIHUHQW YDOXHV RI ] \ R
PAGE 144
a r !! c! X ] + ] fµN R R k ’ k R R W! 9 $ _Y Z R 4 R R 8 ( A ( t im‘ 2 9 R 9 r -/ r )LJXUH $ FRPSDULVRQ RI A YV [n IRU GLIIHUHQW YDOXHV RI ]A
PAGE 145
L R ;r r k R R f§ R H B r‘ E i f§ k k fµ t k L 2n OE G LG )LJXUH 7KH KRUL]RQWDO GLVSHUVLRQ FRHIILFLHQW D YHUVXV WKH \ GLPHQVLRQOHVV URXJKQHVV OHQJWK IRU WZR YDOXHV RI WKH GRZQZLQG GLVWDQFH
PAGE 146
GLVSHUVLRQ FDOFXODWLRQV $OO WKH FXUYHV DUH SORWWHG LQ )LJXUH f IRU WKH FRPSOHWH UDQJH RI [r VWXGLHG 7KH GRWWHG SRUWLRQV r r RI WKH FXUYHV DUH H[WUDSRODWLRQV 6LPLODU SORWV IRU ¾A DQG FA DUH VKRZQ LQ )LJXUHV f DQG f UHVSHFWLYHO\ &DOFXODWLRQ RI FRQFHQWUDWLRQ E\ WKLV PHWKRG UHTXLUHV SULRU NQRZOHGJH RI WKH URXJKQHVV OHQJWK RI WKH VXUIDFH ] DQG WKH IULFWLRQ R YHORFLW\ X$ ,GHDOO\ WKHVH WZR SDUDPHWHUV VKRXOG EH HYDOXDWHG IURP D FXUYHILW RI WKH PHDVXUHG YHORFLW\ SURILOH FORVH WR WKH JURXQG WR WKH YHORFLW\ SURILOH HTXDWLRQ (T f RU ff +RZHYHU WKH YHORFLW\ DW WZR YDOXHV RI ] FDQ DOWHUQDWLYHO\ EH XVHG WR HVWLPDWH ]T DQG X A %XW LI WKH YHORFLW\ LV PHDVXUHG DW RQO\ RQH SRLQW WKH YDOXH RI WKH URXJKQHVV OHQJWK FDQ EH DSSUR[LPDWHG XVLQJ 7DEOH f 1DZURFNL DQG 3DSD 0RQLQ DQG 2EXNKRY f DQG WKH IULFWLRQ YHORFLW\ SURILOH HTXDWLRQ ,Q &KDSWHU WKH GLIIXVLYLW\ ZDV H[SUHVVHG LQ WHUPV RI WKH YDULDQFH RI WKH GLVSHUVLRQ DV . [ f ZKHUH W LV WKH WLPH RI GLIIXVLRQ )URP WKH VLPXODWLRQV WKH FRQFHQn WUDWLRQ GLVWULEXWLRQ LQ WKH \GLUHFWLRQ LV REWDLQHG DW GLIIHUHQW ] YDOXHV :KHQ ORJ FRQFHQWUDWLRQf LV SORWWHG DJDLQVW WKH GRZQZLQG GLVWDQFHf WKH VORSH RI WKH OLQH LV Af 7KLV JUDSK ZDV SORWWHG IRU VHYHUDO UXQV DQG LQ DOO WKH FDVHV VHWV RI DSSUR[LPDWHO\ SDUDOOHO OLQHV ZHUH REWDLQHG )LJ f 7KXV WKH YDULDWLRQ RI ¾ ZLWK ] FDQ
PAGE 147
)LJXUH 7KH KRUL]RQWDO GLVSHUVLRQ FRHIILFLHQW YHUVXV WKH GRZQn ZLQG GLVWDQFH ZLWK WKH GLPHQVLRQOHVV URXJKQHVV OHQJWK DV WKH SDUDPHWHU
PAGE 148
)LJXUH a 7KH YHUWLFDO GLVSHUVLRQ FRHIILFLHQW YHUVXV WKH GRZQZLQG GLVWDQFH ZLWK WKH GLPHQVLRQOHVV URXJKQHVV OHQJWK DV WKH SDUDPHWHU
PAGE 149
)LJXUH 7KH FRQFHQWUDWLRQ FRHIILFLHQW FA YHUVXV WKH GRZQZLQG GLVWDQFH ZLWK WKH GLPHQVLRQOHVV URXJKQHVV OHQJWK DV WKH SDUDPHWHU
PAGE 150
ZLWK WKH KHLJKW DV WKH SDUDPHWHU 7KH VORSH UHSUHVHQWV D\rf
PAGE 151
7DEOH 7KH URXJKQHVV OHQJWKV IRU YDULRXV VXUIDFHV 6XUIDFH 5RXJKQHVV OHQJWK FP 6PRRWK PXG ILHOGV 6PRRWK VQRZ 6XQ EDNHG DOOXYLXP 6DQG 0RZQ *UDVV FPf 6QRZ FPf 0RZQ JUDVV FPf $LUSRUW )DOORZ ODQG 6WHSSH 6KRUW JUDVV ODQG /RQJ JUDVV ODQG :KHDW ILHOG /RQJ JUDVV FPf 6XJDU EHHW SODQWDWLRQ 7XUQLS ILHOG 6RXUFH 1DZURFNL DQG 3DSD 0RQLQ DQG 2EXNKRY
PAGE 152
EH QHJOHFWHG 7KHUHIRUH r\ FDQ EH H[SUHVVHG DV . \ D \ GD = GW f 7KH DYHUDJH WLPH WKH SDUWLFOHV WDNH WR UHDFK D FHUWDLQ SRLQW LV D IXQFWLRQ RI WKH KHLJKW EHFDXVH RI WKH YHORFLW\ SURILOH &RQVHn TXHQWO\ WKH GLIIXVLYLW\ LV D IXQFWLRQ RI WKH KHLJKW 7KH WLPH GW FDQ WKHQ EH WDNHQ DV GW G[ X]f f DQG WKHUHIRUH . \ X D \ GD = G[ f 6LQFH WKH DYHUDJH YHORFLW\ E\ ZKLFK DOO WKH SDUWLFOHV UHDFK D FHUDLQ GRZQ ZLQG GLVWDQFH LV WKH FRQYHFWLRQ YHORFLW\ 8 WKH DYHUDJH U GLIIXVLYLW\ FDQ EH H[SUHVVHG DV . ! 8B \ I D \ f 6LPLODUO\ IRU WKH ]GLUHFWLRQ ZH KDYH WKH IROORZLQJ GD . 8 D ] ] ] G[ f
PAGE 153
DQG . ! ] D ] f 7KHVH WZR GLIIXVLYLWLHV LQ WKH GLPHQVLRQOHVV IRUPV DUH r . \ . ! \ 9 r 8) GD r \B r G[ f DQG r . ] . ! r f§ f§ X X + ) f 1RWH WKDW . DQG .r DUH UHFLSURFDO 3HFOHW QXPEHUV \ t t 7KH GLIIXVLYLWLHV DQG DUH SORWWHG DJDLQVW WKH GLPHQVLRQ fµN r OHVV GRZQZLQG GLVWDQFH [ ZLWK ]A UHVSHFWLYHO\ DV WKH SDUDPHWHU LQ )LJXUHV f DQG f 7KHVH UHVXOWV RIIHU D FKRLFH LQ WKH XVH RI PRGHOV 7KH GLIIXn VLYLWLHV FDQ EH XVHG WR VROYH WKH GLIIXVLRQ HTXDWLRQ RU DOWHUQDWLYHO\ WKH GLVSHUVLRQ FRHIILFLHQWV FDQ EH XVHG WR GLUHFWO\ FDOFXODWH FRQFHQn WUDWLRQV 8VXDOO\ WKH JURXQG OHYHO FRQFHQWUDWLRQV DW \ FAT DUH QHHGHG VLQFH WKHVH FRQFHQWUDWLRQV DUH WKH PD[LPXP SROOXWDQW FRQFHQWUDn WLRQV DW HDFK [
PAGE 154
)LJXUH 7KH GLPHQVLRQOHVV KRUL]RQWDO PDVV HGG\ GLIIXVLYLW\ YHUn VXV WKH GRZQZLQG GLVWDQFH ZLWK WKH GLPHQVLRQOHVV URXJKn QHVV OHQJWK DV WKH SDUDPHWHU
PAGE 155
r 1 )LJXUH 7KH GLPHQVLRQOHVV YHUWLFDO PDVV HGG\ GLIIXVLYLW\ YHUVXV WKH GRZQZLQG GLVWDQFH ZLWK GLPHQVLRQOHVV URXJKQHVV OHQJWK DV WKH SDUDPHWHU
PAGE 156
7KH GLPHQVLRQOHVV JURXQG OHYHO FRQFHQWUDWLRQ F LV REWDLQHG JR IURP (T f E\ OHWWLQJ \ ] DV IROORZV f r r r 7KH JHQHUDOL]HG FKDUWV IRU FA DQG YV [ FDQ EH XVHG WR N HYDOXDWH F 7KH UHVXOWV DUH VKRZQ LQ )LJ f JR &RQFHQWUDWLRQ 'LVWULEXWLRQ LQ WKH 9HUWLFDO 'LUHFWLRQ ,Q 6HFWLRQ f WKH FRQFHQWUDWLRQ GLVWULEXWLRQV LQ ERWK WKH \ DQG ]GLUHFWLRQV ZHUH DVVXPHG WR EH *DXVVLDQ ,Q WKH \GLUHFWLRQ H[SHULPHQWDO UHVXOWV IURP 3URMHFW 3UDLULH *UDVV %DUDG f LQGLFDWH WKDW WKH GLVWULEXWLRQ LV DFWXDOO\ *DXVVLDQ DV LW VKRXOG EH EHFDXVH RI FRPSOHWH KRPRJHQHLW\ DQG SHUIHFW V\PPHWU\ DORQJ WKH D[LV +RZHYHU RQH PD\ H[SHFW GHYLDWLRQV IURP WKH *DXVVLDQ EHKDYLRU LQ WKH YHUWLFDO GLUHFWLRQ EHFDXVH RI QRQKRPRJHQHLW\ 7KXV RQH PLJKW H[SHFW WKH FRQn FHQWUDWLRQ GLVWULEXWLRQ FDQ LQ JHQHUDO EH H[SUHVVHG LQ DQ DQDORJRXV PDQQHU WR (T f E\ UHSODFLQJ WKH *DXVVLDQ GLVWULEXWLRQ IRU WKH ] GHSHQGHQF\ ZLWK DQ DUELWUDU\ IXQFWLRQ )][]f DV IROORZV F[\]f FT[f (\ [\]f )][]f f ZKHUH ( \ H[S f
PAGE 158
(YHQ WKRXJK WKH GLVWULEXWLRQ LV *DXVVLDQ LQ WKH \GLUHFWLRQ WKH VWDQGDUG GHYLDWLRQ R LV ILUVW SHUPLWWHG WR EH D IXQFWLRQ RI WKH \ KHLJKW 7KH DERYH HTXDWLRQ GRHV DVVXPH WKDW )A LV QRW D IXQFWLRQ RI \ DV VHHPV FRQVLVWHQW ZLWK WKH DVVXPSWLRQ RI FRPSOHWH KRPRJHQHLW\ LQ WKH \GLUHFWLRQ 1RZ LW FDQ EH VKRZQ E\ VXEVWLWXWLRQ LQWR WKH GLIIXVLRQ HTXDWLRQ WKDW LI ) LV LQGHSHQGHQW RI \ ( PXVW EH LQGH ] \ SHQGHQW RI ] 7KHUHIRUH ZH FDQ ZULWH ! 2 [ RQO\f f \ \ 7KH LQWHJUDO PDVV EDODQFH DW DQ\ [ FDQ EH ZULWWHQ DV RR RR 4 FX G\ G] )URP WKH GHILQLWLRQ RI D *DXVVLDQ GLVWULEXWLRQ ZH KDYH f ( G\ WW D \ ` f ,Q 6HFWLRQ f WKH FRQYHFWLRQ YHORFLW\ 8 ZDV GHILQHG DV U RR RR FX G\ G] rR 3 22 &2 F G\ G] rr ,7 f
PAGE 159
&RPELQLQJ (TXDWLRQ f f f DQG f DQG VLPSOLI\LQJ ZH REWDLQ F nR WW R R! 8 \ ] ) f ZKHUH ! LV GHILQHG E\ WKH HTXDWLRQ ] > ) G] UU D ] ] f $VVXPLQJ WKDW WKH PROHFXODU WUDQVSRUW FDQ EH QHJOHFWHG LQ WKH DWPRVSKHUH DQG WKDW WKH HGG\ GLIIXVLYLW\ LQ WKH \GLUHFWLRQ .A LV QRW D IXQFWLRQ RI \ WKH GLIIXVLRQ HTXDWLRQ XQGHU VWHDG\ VWDWH FRQGLn WLRQV FDQ EH ZULWWHQ DV ž ,V . _I I . _e ! G[ \ G] ] G] \ 7KH LQGLYLGXDO WHUPV LQ WKH DERYH HTXDWLRQ FDQ EH HYDOXDWHG IURP (TXDWLRQV f DQG f DQG IURP GHILQLWLRQV RI WKH HGG\ GLI IXVLYLWLHV GD . X D UA f \ \ G[ GD ! . XD ! f§ f ] ] G[
PAGE 160
7KHQ WKH GLIIXVLRQ HTXDWLRQ EHFRPHV X ) ] [ . ) ] . ) f§ F . f§r f§r \ -/ r B= ] ] ] F [ R D \ f 6XEVWLWXWLQJ IRU FA IURP (T f ZH KDYH X ( ] [ . ] ) ] X) ] G ! ] X) ] GD fü 3 ] ] ! G[ D ! G[ 8B -n] G[ ] \ ) ) . aA f UUA 7KH ERXQGDU\ FRQGLWLRQV DUH Lf 7KHUH LV QR DFFXPXODWLRQ RU GHSOHWLRQ RI SROOXWDQWV DW WKH JURXQG DQG WKHUHIRUH WKH PDVV IOX[ LQ WKH YHUWLFDO GLUHFWLRQ LV ]HUR RU ) ] ] ] f LLf 7KH FRQFHQWUDWLRQV DW LQILQLWH GLVWDQFHV DUH ]HUR RU ) ] f ] rr
PAGE 161
(TXDWLRQ f DORQJ ZLWK (TXDWLRQV f f f DQG f FRQVWLWXWH D FORVHG V\VWHP RI HTXDWLRQV IRU ) $ ULJRUn RXV DQDO\WLFDO VROXWLRQ LV YHU\ GLIILFXOW WR REWDLQ EXW YDULRXV DSSUR[LPDWH IRUPV RI WKH DQDO\WLFDO VROXWLRQ DUH DV IROORZV $GGLWLYH IXQFWLRQ IRU GHYLDWLRQ IURP *DXVVLDQ EHKDYLRU /HW ) ( ! + = = = f ZKHUH (A! UHSUHVHQWV WKH *DXVVLDQ GLVWULEXWLRQ ZLWK WKH VWDQGDUG GHYLDWLRQ ! ] ( ! ] ]+f \ 9 H[S D ! = ]+f D ! f DQG WKH IXQFWLRQ + UHSUHVHQWV WKH GHYLDWLRQ IURP D *DXVVLDQ VROXWLRQ 6XEVWLWXWLQJ (T f LQWR f ZH KDYH ( ! G] + G] WW D ! ] ] ] f 6LQFH ( UHSUHVHQWV WKH *DXVVLDQ GLVWULEXWLRQ ZH KDYH ] , ( ! G] WW D ! ] ] f
PAGE 162
7KHUHIRUH +] G] f 7KH UHIOHFWLRQ WHUP LQ WKH *DXVVLDQ VROXWLRQ (T f ZDV REWDLQHG E\ WKH DGGLWLRQ RI DQ LPDJLQDU\ VRXUFH DW ] + 8VLQJ D VLPLODU DSSURDFK ZH FDQ ZULWH + K]f Kn]f f ] 8VLQJ WKH ERXQGDU\ FRQGLWLRQ IRU ]HUR PDVV IOX[ DW WKH JURXQG ZH REWDLQ Kn]f K]f f 7KHUHIRUH + K]f K]f f ] 7KH IXQFWLRQV ) DQG + ZHUH HYDOXDWHG IRU WKH UXQ ZLWK ] ] ]r 6HH )LJV f DQG ff 7KH *DXVVLDQ IXQFWLRQ ( ! R ] ZDV REWDLQHG E\ D OHDVW VTXDUH ILW IRU )
PAGE 163
)LJXUH 7KH FRQFHQWUDWLRQ GLVWULEXWLRQ LQ WKH YHUWLFDO GLUHFWLRQ )] WKH *DXVVLDQ IXQFWLRQ (=! DQG WKH GHYLDWLRQ +] DW [r
PAGE 164
)LJXUH 7KH FRQFHQWUDWLRQ GLVWULEXWLRQ LQ WKH YHUWLFDO GLUHFWLRQ )] WKH *DXVVLDQ IXQFWLRQ (=! DQG WKH GHYLDWLRQ +] DW [
PAGE 165
eQ>eQ, f tQF Â772 f@ )LJXUH 1RQ*DXVVLDQ H[SRQHQWLDO IXQFWLRQ IRU YHUWLFDO FRQFHQWUDWLRQ GLVWULEXWLRQ DW [r
PAGE 166
> 'L/=I RfXA 'XAMT )LJXUH 1RQ*DXVVLDQ H[SRQHQWLDO IXQFWLRQ IRU YHUWLFDO FRQFHQWUDWLRQ GLVWULEXWLRQ DW [r
PAGE 167
ZKHUH WKH UHIOHFWLRQ WHUP LV UHSUHVHQWHG E\ D PLUURU LPDJH VRXUFH
PAGE 168
&+$37(5 &21&/86,216 $1' 5(&200(1'$7,216 &RQFOXVLRQV 7KH .LUPVH)DKLHQ VWRFKDVWLF PRGHO RI WXUEXOHQFH ZDV VWXGLHG WKHRUHWLFDOO\ LQ WKLV ZRUN DQG QXPHULFDO VLPXODWLRQ RI WKH IORZ LQ WKH DWPRVSKHUH ZDV FDUULHG RXW RQ D GLJLWDO FRPSXWHU 7KH PRGHO DVVXPHV WKDW WKH GLVSODFHPHQW RI D IOXLG SDUWLFOH LV D 0DUNRY SURFHVV DQG WKXV REH\V D )RNNHU3ODQFN HTXDWLRQ 7KH )RNNHU 3ODQFN HTXDWLRQ ZDV FRPSDUHG ZLWK WKH GLIIXVLRQ HTXDWLRQ DQG LW ZDV VKRZQ WKDW WKH GHWHUPLQLVWLF YHORFLW\ XVHG LQ WKH )RNNHU3ODQFN HTXDn WLRQ LQ JHQHUDO LV QRW HTXDO WR WKH PHDQ ZLQG YHORFLW\ DQG WKDW WKH /DJUDQJLDQ GLIIXVLYLW\ LV HTXDO WR WKH (XOHULDQ GLIIXVLYLW\ RQO\ XQGHU FHUWDLQ FLUFXPVWDQFHV $Q HTXDWLRQ IRU WKH PHDQ VSUHDG RI WKH SOXPH ZDV GHULYHG DVn VXPLQJ WKH SDUWLFOH PRWLRQ IROORZHG /DJUDQJLDQ VLPLODULW\ WKHRU\ DQG WKDW WKH DXWRFRUUHODWLRQ FRHIILFLHQW ZDV RI WKH H[SRQHQWLDO GHFD\ W\SH W\SLFDO RI D 0DUNRY SURFHVV 7KH YHORFLW\ DQG WHPSHUDWXUH SURILOHV IRU QHXWUDO DQG QRQn QHXWUDO FRQGLWLRQV DUH GLVFXVVHG $ PHWKRG IRU FDOFXODWLQJ WKH YDULn DQFH RI WKH /DJUDQJLDQ DFFHOHUDWLRQV DQG WKH PHDQ WLPH VWHSV IURP WKH WXUEXOHQW HQHUJ\ GLVVLSDWLRQ UDWH LV SUHVHQWHG 6WDELOLW\ SDUDPHWHUV DUH GHILQHG LQ WHUPV RI WKH SUHVVXUH JUDGLHQW WKH JUDGLHQW VLPLODULW\ OHQJWK DQG WHPSHUDWXUH VFDOHV WKH WHPSHUDWXUH DW WKH JURXQG DQG WKH
PAGE 169
DGLDEDWLF ODSVH UDWH 7KH PHDQ /DJUDQJLDQ DFFHOHUDWLRQ LV H[SUHVVHG LQ WHUPV RI WKHVH VWDELOLW\ SDUDPHWHUV DQG WKHLU YDULDWLRQ ZLWK KHLJKW LV VKRZQ 7KH GLVWULEXWLRQ FXUYHV ZHUH GLVFUHWL]HG DQG D VHQVLWLYLW\ DQDO\VLV ZDV FDUULHG RXW IRU WKH WRWDO QXPEHU RI SDUWLFOHV UHOHDVHG WKH VDPSOLQJ WLPH DQG WKH FHOO VL]H 7KH PRGHO LV VKRZQ WR SUHGLFW WKH H[SHULPHQWDO FRQFHQWUDWLRQ GLVWULEXWLRQ FXUYHV UHDVRQDEO\ ZHOO IRU GLIIHUHQW VWDELOLW\ FRQGLWLRQV 7KH UHVXOWV LQ ERWK WKH \ DQG ]GLUHFWLRQV IRU GLIIHUHQW VWDELOLW\ FRQ GLWLRQV DQG GRZQZLQG GLVWDQFHV DUH FRPSDUHG ZLWK WKH H[SHULPHQWDO DWn PRVSKHULF GDWD RI 3URMHFW 3UDLULH *UDVV DQG ZLWK SUHGLFWLRQV PDGH E\ RWKHU PHWKRGV 7KH UHVXOWV DUH SUHVHQWHG LQ D XVHIXO IRUP IRU QHXWUDO FRQGLn WLRQV 7KH KRUL]RQWDO DQG YHUWLFDO GLVSHUVLRQ FRHIILFLHQWV DQG WKH HGG\ GLIIXVLYLWLHV DUH SORWWHG DJDLQVW WKH GRZQZLQG GLVWDQFH IRU YDULn RXV YDOXHV RI WKH GLPHQVLRQOHVV URXJKQHVV SDUDPHWHU 7KH UHVXOWV RIIHU D FKRLFH LQ WKH XVH RI PRGHOV RQH FDQ HLWKHU VROYH WKH GLIIXn VLRQ HTXDWLRQ LQ WKH GLIIHUHQWLDO IRUP XVLQJ WKH FRUUHODWHG HGG\ GLIIX VLYLWLHV RU KH FDQ XVH WKH LQWHJUDWHG IRUP ZLWK WKH FRUUHODWHG GLVSHUn VLRQ FRHIILFLHQWV $ PRUH ULJRURXV DSSURDFK WR WKH FRQFHQWUDWLRQ GLVWULEXWLRQ LQ WKH YHUWLFDO GLUHFWLRQ LV SUHVHQWHG DQG LWV DSSUR[LPDWH IRUPV DUH GLV FXVVHG
PAGE 170
5HFRPPHQGDWLRQV )RU VWHDG\ VWDWH FRQGLWLRQV WKH REYLRXV H[WHQVLRQV RI WKH VWXG\ DUH VLPXODWLRQV IRU OLQH VRXUFHV HJ KLJKZD\Vf DQG DUHD VRXUFHV 7KH SDUWLFOHV ZRXOG EH UHOHDVHG IURP GLIIHUHQW SRLQWV DQG WKH DYHUDJHV RI WKH UHOHDVH SRLQWV ZRXOG UHSUHVHQW WKH VRXUFHV $ PRUH XVHIXO DSSOLFDWLRQ ZRXOG EH WKH FDVH RI UHDFWLYH SROOXWDQWV ZLWK WKH LQFOXVLRQ RI D UHDFWLRQ WHUP LQ WKH VLPXODWLRQ )RU D VLPn SOH UHDFWLRQ WKLV FRXOG EH HDVLO\ GRQH E\ H[SUHVVLQJ WKH PDVV GHn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n WLRQV RIWHQ HQFRXQWHUHG LQ UHDO OLIH 7KHVH WZR FDVHV UHSUHVHQW YDU\LQJ VRXUFH VWUHQJWKV DQG VKRXOG EH VWXGLHG
PAGE 171
6LQFH ZHDWKHU FRQGLWLRQV DUH QHYHU LQ D VWHDG\ VWDWH WKH HIIHFW RI YDU\LQJ WKH VWDELOLW\ FRQGLWLRQV ZLWK WLPH VKRXOG EH VWXGLHG )URP DQ DFDGHPLF UHVHDUFK SRLQW RI YLHZ WKH PRGHO FRXOG EH UHn ILQHG ZLWK UHJDUG WR LWV VWDWLVWLFDO EHKDYLRU $ PDVV GLVWULEXn WLRQ RI SDUWLFOHV PD\ EH LQWURGXFHG RU DOWHUQDWLYHO\ WKH VLPXODn WLRQ FRXOG EH FDUULHG RXW LQ WKH ZDYH QXPEHU VSDFH 7KH PRGHO FRXOG EH PRGLILHG WR SUHGLFW NLQHWLF EHKDYLRU RI FDWDn O\WLF SDUWLFOHV 7KH SUREDELOLW\ GLVWULEXWLRQV IRU WKH DUULYDO DGVRUSWLRQ DQG UHDFWLRQ RI WKH IOXLG SDUWLFOHV FRXOG EH LQFRUn SRUDWHG LQ WKH PRGHO $ IL[HG EHG FROXPQ FDQ EH ORRNHG XSRQ DV DQ DQDORJRXV VLWXDWLRQ WR D WXUEXOHQW IORZ ZLWK IL[HG HGGLHV )ORZ EHKDYLRU LQ D FROXPQ FRXOG EH VWXGLHG WR DLG LQ WKH GHVLJQ RI VXFK FROXPQV
PAGE 172
$33(1',; $ ',6&5(7,=$7,21 2) ',675,%87,21 &859(6 $1' 7+( 0217( &$5/2 6,08/$7,21 0(7+2' $ SDUWLFOH GLVSODFHPHQW GXULQJ D WLPH LQWHUYDO $WA LV JLYHQ E\ (T f ZKLFK LV UHSHDWHG KHUH $W $W $; 9 $W D f§ D U O P [ L P P P P $Of ZKHUH $;BA LV WKH SDUWLFOH GLVSODFHPHQW LQ WKH Lf§ GLUHFWLRQ GXULQJ P WKH WLPH LQWHUYDO $W 9 D DQG D? DUH WKH WRWDO YHORFLW\ Pf¬ L L L P P P PHDQ /DJUDQJLDQ DFFHOHUDWLRQ DQG WKH IOXFWXDWLQJ /DJUDQJLDQ DFFHOHUDn WLRQ DW WKH EHJLQQLQJ RI WKH LQWHUYDO $WA 7KH SDUWLFOH DW D JLYHQ LQVWDQW RI WLPH PD\ SRVVHVV DQ\ UDQGRP YDOXHV RI WKH YHORFLW\ DFFHOn HUDWLRQ DQG WLPH VWHS %XW WKHVH TXDQWLWLHV DUH VWRFKDVWLF LQ QDWXUH DQG REH\ FHUWDLQ SUREDELOLW\ GLVWULEXWLRQ ODZV (DFK FRPSRQHQW RI WKH IOXFWXDWLQJ YHORFLW\ 9A DQG RI WKH DFFHOHUDWLRQ DA LV GLVWULEXWHG QRUPDOO\ ZLWK WKH VWDQGDUG GHYLDWLRQ & DQG 2 UHVSHFWLYHO\ %\ YL DL GHILQLWLRQ WKH PHDQ RI D IOXFWXDWLQJ TXDQWLW\ LV ]HUR 7KH WLPH VWHSV DUH GLVWULEXWHG H[SRQHQWLDOO\ ZLWK D PHDQ RI $ LQ WKH Lf§ GLUHFWLRQ L )RU WKH VLPXODWLRQ LW LV QHFHVVDU\ WR GLJLWL]H WKH GLVWULEXWLRQ IXQFWLRQV /HW XV FRQVLGHU WKH YHORFLW\ FRPSRQHQW Xn 7KH SUREDELOLW\ GHQVLW\ RI WKH YDULDEOH Xn LV JLYHQ E\
PAGE 173
-/ IXnf WWJAf H[S $f $OVR WKH SUREDELOLW\ GHQVLW\ RI D YDULDEOH V ZLWK D PHDQ RI ]HUR DQG D YDULDQFH RI RQH LV JLYHQ E\ $f 7KH IOXFWXDWLQJ YHORFLW\ X7 LV WKHQ UHODWHG WR V DV X * V $f X $OO WKH IOXFWXDWLQJ YHORFLW\ DQG DFFHOHUDWLRQ FRPSRQHQWV DUH JHQHUDWHG IURP WKH VWDQGDUG YDULDEOH V WKURXJK WKH XVH RI HTXDWLRQV VLPLODU WR (T $f 7KH FXPXODWLYH GLVWULEXWLRQ IXQFWLRQ )Vf IRU V PD\ EH ZULWWHQ DV V )Vf ‘ I H[S Vnf GVn $f WW 7KH GLVFUHWL]DWLRQ LV DFKLHYHG E\ GLYLGLQJ WKH GLVWULEXWLRQ IXQFWLRQ )Vf LQWR HTXDO SDUWV 6HH )LJ $Of $ XQLIRUPO\ GLVn WULEXWHG UDQGRP QXPEHU LV JHQHUDWHG WR REWDLQ WKH SDUWLWLRQ QXPEHU RQ WKH )Vf FXUYH 7KH ZHLJKWHG DYHUDJH YDOXH RI WKH YDULDEOH V LQ
PAGE 174
b )LJXUH $O 'LVFUHWL]DWLRQ RI D *DXVVLDQ FXPXODWLYH GLVWULEXWLRQ IXQFWLRQ
PAGE 175
WKLV LQWHUYDO LV WKH QRUPDOO\ GLVWULEXWHG UDQGRP QXPEHU V 7KH ZHLJKWHG DYHUDJH YDOXH RI V LQ WKH LUA SDUWLWLRQ VDYJf LV JLYHQ E\ Q VI VfGV V f Q DYJ f§ Q V $f Q IVfGV Q ZKHUH VA DQG VQA DUH WKH YDOXHV RI V DW WKH EHJLQQLQJ DQG DW WKH HQG RI WKH LUA LQWHUYDO UHVSHFWLYHO\ $Q ,%0 GLJLWDO FRPSXWHU ZDV XVHG LQ WKLV ZRUN 7KH UDQGRP QXPEHU JHQHUDWRU GHYHORSHG E\ ,QWHUQDWLRQDO %XVLQHVV 0DFKLQHV f ZDV XVHG WR REWDLQ WKH XQLIRUPO\ GLVWULEXWHG UDQGRP QXPEHUV 2QO\ WKH SRVLWLYH KDOI RI WKH GLVWULEXWLRQ FXUYH )Vf LV XVHG WR REWDLQ WKH PDJQLWXGHV RI WKH DFFHOHUDWLRQ IOXFWXDWLRQV 7KH VLJQ LV REWDLQHG UDQGRPO\ DQG LQGHSHQGHQWO\ E\ VDPSOLQJ D YHORFLW\ 9 WHVW IURP WKH YHORFLW\ GLVWULEXWLRQ FXUYH 7KH FULWHULRQ XVHG IRU GHWHUn PLQLQJ WKH VLJQ RI WKH DFFHOHUDWLRQ LV WKH IROORZLQJ ,I 9 ! 9 L f§ L WHVW WKHQ D ! [ D L RWKHUZLVH
PAGE 176
ZKHUH LV WKH IOXFWXDWLQJ YHORFLW\ RI WKH SDUWLFOH LQ WKH Lf§ GLUHFWLRQ DW WKH WLPH RI WKH VDPSOLQJ $ VLPLODU PHWKRG LV XVHG IRU REWDLQLQJ WKH UDQGRP YDOXHV RI WKH WLPH VWHSV 7KH SUREDELOLW\ GHQVLW\ IRU WKH WLPH VWHSV PD\ EH H[SUHVVHG DV $f $W ! L ZKHUH ; [V WKH PHDQ RI WKH YDULDEOH $W 7KH SUREDELOLW\ GHQVLW\ W L L RI WKH YDULDEOH U ZLWK D PHDQ RI RQH LV JLYHQ E\ I Uf H[S Uf $f $QG WKH FXPXODWLYH GLVWULEXWLRQ FDQ EH ZULWWHQ DV )Uf H[S Uf $f 7KH GLVWULEXWLRQ FXUYH )Uf )LJ $f LV XVHG WR REWDLQ WKH UDQGRP YDOXHV RI WKH H[SRQHQWLDOO\ GLVWULEXWHG VWDQGDUG YDULDEOH U DQG WKH WLPH VWHSV DUH REWDLQHG WKURXJK WKH XVH RI WKH IROORZLQJ UHODWLRQ
PAGE 177
)Uf U DYJAQ U )LJXUH $ 'LVFUHWL]DWLRQ RI DQ H[SRQHQWLDO FXPXODWLYH GLVWULEXWLRQ IXQFWLRQ
PAGE 178
$W U $ L W L $f 7KH SDUWLWLRQ QXPEHU RQ WKH )Uf FXUYH LV REWDLQHG IURP WKH XQLIRUP GLVWULEXWLRQ UDQGRP QXPEHU JHQHUDWRU
PAGE 179
$33(1',; % &20387(5 352*5$06 $OO WKH SURJUDPPLQJ ZDV GRQH LQ WKH )RUWUDQ ,9 ODQJXDJH 7KH IROORZLQJ IRXU PDLQ SURJUDPV ZHUH ZULWWHQ IRU WKH 0RQWH &DUOR VLPXODWLRQ RI WKH PRGHO Lf352% 7KH SURJUDP GLVFUHWL]HV WKH GLVWULEXWLRQ FXUYHV DQG FDOFXODWHV WKH QRUPDOL]LQJ FRQVWDQWV XVHG WR REWDLQ WKH SDUWLWLRQ QXPEHUV RQ WKH GLVWULEXWLRQ FXUYHV IURP WKH UDQGRP QXPEHUV JHQHUDWHG E\ WKH UDQGRP QXPEHU JHQn HUDWRU LLf35),/6 7KLV SURJUDP HYDOXDWHV WKH IULFWLRQ YHORFLW\ WKH JUDGLHQW VLPLODULW\ OHQJWK DQG WHPSHUDWXUH VFDOHV DQG WKH YHORFLW\ SURILOH H[SRQHQW LLLf0217( &$5/2 6,08/$7,21 7KLV SURJUDP UHDGV LQ DOO WKH QHFHVVDU\ GDWD IURP WKH FDUGV FDUULHV RXW WKH FHOO VL]H FDOFXODWLRQV VLPXn ODWHV WKH SDUWLFOH SDWKV VDPSOHV WKH SDUWLFOH SRVLWLRQV DQG FDOFXODWHV WKH FRQFHQWUDWLRQV LYf6,*0$6 7KH SURJUDP UHDGV WKH VXPPHG YDOXHV RI WKH FRQFHQn WUDWLRQV LQ WKH \ DQG ]GLUHFWLRQV GLVFUHWL]HV WKH YHORFLW\
PAGE 180
SURILOH FKHFNV WKH LQWHJUDO PDVV EDODQFH DQG FDOFXn ODWHV WKH KRUL]RQWDO DQG YHUWLFDO GLVSHUVLRQ FRHIIL fµ FLHQWV DQG WKH HGG\ GLIIXVLYLWLHV 7KH IROORZLQJ VXEURXWLQHV DUH WKRVH RI 'U $: :HVWHUEXUJ 'HSDUWPHQW RI &KHPLFDO (QJLQHHULQJ 8QLYHUVLW\ RI )ORULGD Lf (0,164 ,W LV D OHDVW VTXDUH FXUYHILW SURJUDP WR HYDOXDWH RQH RU PRUH SDUDPHWHUV LQ D JLYHQ HTXDWLRQ LLf 548$' 7KLV SURJUDP HYDOXDWHV WKH PLQLPD RI D JLYHQ IXQFn WLRQ 7KH SURJUDP ZDV XVHG DV D OHDVW VTXDUH FXUYHILW IRU RQH SDUDPHWHU 7KH IROORZLQJ VXEURXWLQHV DUH DYDLODEOH IURP WKH ,%0 6FLHQWLILF 6XEURXWLQH 3DFNDJH ,QWHUQDWLRQDO %XVLQHVV 0DFKLQHV f Lf 6* ,W LV D VPRRWKLQJ VXEURXWLQH WR PLQLPL]H WKH VFDWWHU LQ GDWD WKDW DUH DYDLODEOH LQ D WDEXODWHG IRUP LLf '*7 7KLV VXEURXWLQH HYDOXDWHV WKH GHULYDWLYHV RI D IXQFn WLRQ ZKHQ WKH GDWD DUH DYDLODEOH LQ D WDEXODWHG IRUP
PAGE 181
&rrrrrrrrrU rrrrrrr‘rtrb & rrrrr 354% rrrrr & 385326( & ',6&5(7,=$7,21 2) ',675,%87,21 &859(6 & Of *$866,$1 :,7+ 0($1 } $1' 9$5,$1&(6 & f (;321(17,$/ :,7+ 0($1 & 120(1&/$785( & ) &808/$7,9( ',675,%87,21 )81&7,21 2) /(1*7+ & 131 25 13( & 3131 rr3131 180%(5 2) 3$57,7,216 2) 1250$/ ) &859( & (36) $//2:$%/( (5525 ,1 ) & '; ),567 75,$/ ,1&5(0(17 2) ; & 131 180%(5 2) 3$57,7,216 2) 1250$/ ) &859( & 331 1250$/,=,1* )$&725 )25 1250$/ &859( & ')1 21( ,1&5(0(17 2) ) & ;6$9 1250$//< ',675,%87(' 9$/8(6 2) 7+( 9$5,$%/( ; & 313(13(33(')( ( 5()(56 72 (;321(17,$/ &859(6 & 6(( 3131r 131} 331 $1' ')1f & <6$9 (;321(17,$//< ',675,%87(' 9$/8(6 2) 7+( & 9$5,$%/( < & ,51' 5$1'20 180%(5 &rrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrr ,03/,&,7 ,17(*(523 f ',0(16,21 ;6$9f<6$9f &1250$/ ',675,%87,21 5($' 3131(36)'; )250$7,)f 35,17 )250$7;n1250$/ ',675,%87,21 )81&7,21 :,7+ 0($1 $1'n n 6,*0$ f 131 rr3131 331 rr,3131f 35,17 ,,3131131331 )250$7;n32:(5 180%(5 2) 3$57,7,216 2) ) n,; n180%(5 2) 3$57,7,216 2) )fµ; n fµ , ; n1250$/,=,1* )$&725n;fµ n f ;131 131 ')1 ;131 35,17 ')1(36) )250$7;n')1 fµ(;n(36) n(f 131 -; 5 6457 f )3 ')1 6';% $'; '; ;, ; '; $'; ; ';; )& r(5);5f
PAGE 182
,))&/7)3f *2 72 ,)$%6)&)3f/7(36)f *2 72 ; ;'; '; '; *2 72 ';% r(;3r;,r;,f(;3r;r; f f 6';% 6';%';% ;6$9-f ';%')1 ;6$9---f f§; 6 $9 f )3 ')1) 3 --,)-(4131f *2 72 *2 72 ;% 6'; ;6$9-f ';%')1 ;6$9---f ;6$9-f 35,17 )250$7;n1r,;n;6$9f 131 35,17 -;6 $9^f )250$7;;(f & (;321(17,$/ ',675,%87,21 5($' } 313( ) 250 $7f 35,17 )250$7nOn;n(;321(17,$/ ',675,%87,21 )81&7,21 :,7+n n 3$5$0(7(5 f 13( rr313( 33( rr313(f 35,17 313(13(33( ) << 6'<% 13(0 13(f§ ;13( 13( ')( ;13( 35,17 ')( )250$7;r')( n( f '2 1 13 (0, ) )')( & << & $/2*L)f '<% (;3&fr&f(;3&fr & f 6'<% 6'<%'<% <6$91f '<%')( << & 1 13( ) '<% 6'<% <6$91f '<%')(
PAGE 183
35,17 )250$7 ;n 1nr,;}n<6$9nf '2 . 13( 35,17 .<6 $9. f & 1250$//< ',675,%87(' 5$1'20 180%(56 5($' ,51' )250$7f 35,17 )250$7rfµ,}r 1250$/,=(' 5$1'20 180%(561250 ',675,%87('n n 5$1'20 180%(56n f '2 ,51' ,51'r ,),51'f ,51' ,51' 13571 ,51'331, 51'1 ;6$913571f 35,17 1357151'1 )250$7;,;(f & (;321(17,$//< ',675,%87(' 5$1'20 180%(56 35,17 )250$7nnn 1250$/,=(' 5$1'20 180%(56n n(;3 ',675,%87(' 5$1'20 180%(56nf '2 ,51' ,51r ,),51'f ,51' ,51' 1357( ,51'33( 51'( <6$91357(f 35,17 1357(51'( )250$7;;(f 35,17 ,51' )50$7;n,51' nf 6723 (1'
PAGE 184
& rrrrr 35),/6 rrrrr & 385326( & (9$/8$7,21 2) & 6/23( 2) 7+( 87+(7$ /,1( & 7+(7$ $7 7+( *5281' & )5,&7,21 9(/2&,7< & *5$',$(17 6,0,/$5,7< /(1*7+ 6&$/( & *5$',(17 6,0,/$5,7< 7(03(5$785( 6&$/( & 9(/2&,7< 352),/( (;321(17 & 120(1&/$785( F 151 581 180%(5 F 1(;7 ,) 7+,6 ,6 7+( /$67 581 F 1;= ,) 7+( =9$/8(6 +$9( 127 &+$1*(' )520 7+( F 35(9,286 581 F 1358 ,) 7+(5( $5( $'',7,21$/ 89$/8(6 F 1357 ,) 7+(5( $5( $'',7,21$/ 79$/8(6 F 13 180%(5 2) =9$/8(6 )25 87+(7$ /,1( 5($' 21/< F ,) 1;=1(O F = +(,*+7 F 7 7(03(5$785( F 8 9(/2&,7< F 138 727$/ 180%(5 2) 89$/8(6 5($' 21/< ,) 13581(2 F =8 = 9$/8(6 &255(6321',1* 72 $// 79$/8(6 F 13 7 180%(5 2) 7 9$/8(6 F ,7 = 9$/8(6 &255(6321',1* 72 $// 7 9$/8(6 F 77 7(03(5$785( F (/3 *5$',(17 6,0,/$5,7< /(1*7+ 6&$/( F = $%6,(/3f e e r r r r rrr r r err rrrrr rrrr r^r rrrrrrrrrrrrr rrrrrrrrrr & F ',0(16,21 =fW=8f=7f7f77^7+!8f 5($' } 151} 1(;71;= r 13581357 ,)1;=(42f *2 72 5($' 13 5($' =.f . 13f 5($' 7.f . 13f 5($' 8.f . O13f = '2 . 13 7+.f 7.fr=.f 35,17 151 & &$// 87+=7+87137+75'8'7+f %(7 $3 ,)'8'7+/7f %(7$3 ,)1358(42f *2 72 5($' r 138 5($' =8.f . 138f8.f . 138f *2 72
PAGE 185
'2 . 13 =8^.f =.f 138 13 &$// 867=88138=2'8'7+7+75%(7$3867$5W(/3=Of ,)1357(42f *2 72 5($' } 137 5($' r =7.f} . }137f77.f . 137f *2 72 137 13 '2 . } 13 7 =7.f =.Of 7 7. f 7.O` 5($' =7f77f '2 . 137 7+.f 77.fr=7.f 767$5 867$5'8'7+ &$// 7+5^=77+137=2'8'7+%(7$3767$5A=7+75f (/3 867$5r'8'7+r7+75 = $%6r(/3f 35,17 (/3= &$// $/3=88138(/3W=2W=O%(7$3867$5$/3+$f ,)1(;7(4f *2 72 )250$7f )250$76) f )250$7f¬r;fµ 581 180%(5 ,6 n,n nf )50$7,4;rrrrrr /35,0( n(;}fµrrrrr = n (f 6723 (1'
PAGE 186
RRQRRRRRRQR -rrrrrrrrrrrrrrrrrrrrrrrrrrrrr rrrr rrrrrrrrrrrrrrr rrrrrrrrrrrr rrrrr 87+ rrrrr 385326( &$/&8/$7,21 2) 7+( 6/23( 2) 87+(7$ /,1( $1' ),567 9$/8( 2) 7+(7$ $7 7+( *5281' 120(1&/$785( 7+ 7+(7$327(17,$/ 7(03(5$785( 7+75 7+(7$5 7+(7$ $7 7+( *5281' '8'7+ '(5,9$7,9( 2) 8 :,7+ 5(63(&7 72 7+(7$ 7+& &$/&8/$7(' 7+(7$ rrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrr e 68%5287,1( 87+^=7+87137+75'8'7+f ',0(16,21 =f7+f7+&f)f563^f,63f$f ( & f}8f7 Âf}:f & '2 (, f $ ^ f 7+13f7+ff8^13f8f f $f 7+ff§$^f8 f ,7(67 & &$// 50,164,fµ7+78/,1(n13)$(O563,63, 7(67 :f & *2 72 f ,7(67 '2 . 13 ),.f 7+.f$f$fr8.f *2 72 35,17 ,7(67 7+75 $^f '8'7+ $f 35,17 7+75'8'7+ '2 . 13 7+&.f 7+758.f'8'7+ 35,17 =.f8.f7.f7+.f7+&.f )250$7;f¬7+(7$f§5 )520 87+(7$ /,1( n(; r '8'7+ n (;fµ=fµ ; r8n ; fµ7r ;fµ7+(7$r; n&859( ),7 7+(7$rf )50$7;(ff )250$7r ,7(67 rf 5(7851 (1'
PAGE 187
RQQQQRRRRRRR rrrrrrrrrrrrrrrrrrrerrrrrrrrttrrttrttbtb rrrrr 867 rrrrr 385326( &$/&8/$7,21 2) )5,&7,21 9(/2&,7< $1' *5$',(17 6,0,/$5,7< /(1*7+ 6&$/( 120(1&/$785( =2 528*+1(66 /(1*7+ %(7$3 %(7$35,0( 7+( &2167$17 :,7+ 7+( /,1($5 7(50 ,1 7+( /2* /,1($5 352),/( 867$5 )5,&7,21 9(/2&,7< 8& &$/&8/$7(' 8 68%5287,1( 867=813=2'8'7+7+75%(7$3867$5(/3=f ',0(16,21 = f }8f8&f52/f,1$f ( % %(7$3r'8'7+7+75 ,75/ = ,75/ ,75/ 11 '2 . 13 ,)=W.f/(=Lf 11 11 ,),11/7f 11 ,7(67 &$// 548$'Â,7(67&)}( 5/ ,1$f *2 7f ,7(67 ) '2 . 11 ) )8.f&r$/2* =. f = f =2f%r=^.ffrr *2 72 35,17 ,7(67 867$5 &r (/3 &r'8'7+r7+75 = $%6Lr(/3f , ) ,75/(4f *2 72 35,17 867$5(/3 *2 72 35,17 867$5(/3= '2 . 11 8&.f &r$/2*=.f=f=f%r=.f 35,17 =.f8.f8&.f )50$7Â;n$3352;,0$7( 9$/8(6 2) 867$5/35,0( $5(} ; } ) 4 f f )250$7^ 2;nrrrrr 867$5 fµ( 2; n/35, 0( )520 n n867$5 $1' 35(9 7+7m fµ(;m= r(;n=fµ; n8n;n&859( ),7 8 rf )50$7;(ff )250$7n ,7(67rn f 5(7851 (1'
PAGE 188
RRRRRRRR rrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrr rrrrr 7+5 rrrrr 385326( &$/&8/$7,21 2) 7+(7$5 120(1&/$785( 767$5 *5$',(17 6,0,/$5,7< 7(03(5$785( 6&$/( 7+& &$/&8/$7(' 7+(7$ 9$/8(6 rrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrr 68%5287,1( 7+5= r7+13=2'8'7+%(7$3767$5=7+75f ',0(16,21 =f7+f7+&f}52/f,1$^f ( % %(7$3r'8'7+'8'7+ 11 '2 . 13 ,),=.f/(=,f 11 11 ,),11/7f 11 ,7 (6 7 &$// 548$'+,7(677+75)( , 52/,1$ f *2 72 ^}f ,7(67 ) '2 . 11 ) ) 7+.`7+75767$5r$/4*=.f=f=2f%r=.f7+75frr *2 72 35,17 ,7(67 35,17 7+75767$5 '2 . 11 7+&.f 7+75767$5r$/2*=.f=f=2f%r=,.f7+75 35,17 =.f7+.f7+&.f )250$7,O2;nrrrrr 7+(7$5 n(;rrrrrr 767$5 fµ( 2;fµ= r ; n7+(7$r };n&859( ),7 7+(7$rf )250$7;(f f )250$7r ,7(67 fµf 5(7851 (1'
PAGE 189
RRRRRRRRR rrrrrmrerr mmcIHrrrr rr A/3 385326( &$/&8/$7,21 2) 7+( 9(/2&,7< (;321(17 120(1&/$785( $/3+$ 9(/2&,7< (;321(17 3+, ',0(16,21$/(66 9(/2&,7< $7 = =23 =235,0( =2(/3 68%5287,1( $/3=813(/3=2=%(7$3867$5$/3+$f ',0(16,21 =f8 f 8&f52/f,1$^f 11 '2 . 13 ,)=.f*(=Of 11 11 ,),11/7f 11 11 13f§11 3+, $/2* =, =2f=2f=r%(7$3(/3 =23 =2(/3 % 3+,*867$5 ( , 7(67 &$// 548$'. , 7(67 $/3+$ ) ( L52/ ,1$f *2 72 f ,7(67 ) '2 . 11}1 3 ) )8.ff§ %r=.f=frr$/3+$frr *2 72 35,17 ,7(67 35,17 $/3+$3+,=23 '2 . 1113 8&.f %r=.f=frr$/3+$ 35,17 =.f8.f8&.f )50$7Â; rrrrrr $/3+$ n(;nrrrrr 3+ , fµ(; rrrrrr =2f§35,0( r(;n= r;n 8n;r 8& rf )250$7;( f f )250$7n ,7(67 nf 5(7851 (1'
PAGE 190
RQ RRRRRRRR rrrrmrrreerrrrrrerrrrrrrrrerrmrrrererrmemrrrrrerr rrrrr 0217( &$5/2 6,08/$7,21 rrrrr 385326( 72 6,08/$7( 7+( 3$57,&/( 3$7+6 $1' &$/&8/$7( &21&(175$7,216 rrrrrrrrrrrr e e r r r rrrr rr r r Wrr rr r r r r cr r r r — r ` e &20021 $O;6$9,f<6$9f;01f ;0;,f=01f <0;f';f}'
PAGE 191
& rrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrr rrrrrrFreeLrfmerrrrArr & F F F F F F F F F F F F F F F F F F F F F F F F F F F F F F F F F F F F F F F F F F F F F F F rrrrr 75 rrrrr 385326( 72 5($' )25 7+( 120(1&/$785( 135% 7+( 352%$%,/,7< 6,08/$7,21 $1' 785%8/(1&( '$7$ 1(&(66$5< ;6$9 <6$9 7,7/( &207 13 67 +6 180%(5 2) 7+( 3$57,7,216 2) 7+( 352%$%,/,7< ',675,%87,21 &859( ',6&5(7,=(' 9$/8(6 2) 7+( 1250$//< ',675,%87(' 67$1'$5' 9$5,$%/( ',6&5(7,=(' 9$/8(6 2) 7+( (;321(17,$//< ',675,%87(' 67$1'$5' 9$5,$%/( 7,7/( )25 7+( 352*5$0 &+$5$&7(56 /21* $1' &207 &200(176($&+ &200(17 &+$5$&7(56 727$/ 180%(5 2) 3$57,&/(6 5(/($6(' 6$03/,1* 7,0( 67$&. +(,*+7 ;0$;<0$;=0,1=0$; 5(*,21 2) ,17(5(67 ,1 7+( ; < $1' = ',5(&7,216 5(63(&7,9(/< 21&( 7+( 3$57,&/(6 /($9( 7+,6 5(*,21 7+( 1(;7 3$57,&/( ,6 5(/($6(' 4 6285&( 675(1*7+ 6,*86,*96,*: 785%8/(1&( ,17(16,7,(6 ,1 7+( ;< $1' = ',5(&7,216 5(63(&7,9(/< %;%<%= 7+( /$*5$1* , $1 (8/(5,$1 3$5$0(7(5 ,1 7+( ;< $1' = ',5(&7,216 5(63< &$;&$<&$= &2167$176 2) 3523257,21$/,7< )25 7+( $&&(/(5$7,216 ,1 7+( ;< $1' = ',516 5(63< $&&(/(5$7,1 &$r= &/0';&/0'<&/0'= &2167$176 2) 3523257,21$/,7< )25 7+( 0($1 7,0( 67(36 /$0'$7 &/0'r=% &2167$17 9(/2&,7< 5($' 21/< ,) .8%5(4Of &2167$17 0($1 9(57,&$/ /$*5$1*,$1 $&&(/(5$7,21 5($' 21/< ,) .735(4Of 528*+1(66 /(1*7+ 7+( *5$',(17 6,0,/$5,7< /(1*7+ 6&$/( )5,&7,21 9(/2&,7< *5$',(17 6,0,/$5,7< 7(03(5$785( 6&$/( 9(/2&,7< 352),/( (;321(17 5()(5(1&( 7(03(5$785( 5()(5(1&( 35(6685( 9(57,&$/ 35(6685( *5$',(17 180%(5 2) '2:1 :,1' ',67$1&(6 180%(5 2) *5,'6 ,1 7+( < $1' = ',516 5(63< 7+( ;9$/8(6 $7 7+( %(*,11,1* $1' $7 7+( (1' 2) 7+( 6$03/,1* *5 ,'69(&7256 2) 1&; /(1*7+ <0;=01=0; 7+( 5(*,21 (1&/26,1* 7+( 6$03/,1* *5,'6 9(&7256 2) 1&; /(1*7+ ,51' ),567 5$1'20 180%(5 Trrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrr 68%5287,1( '75 8%5& $ = %$5 =2 (/3 867$5 767$5 $/3+$ 75() 635() '3 '= 1&; 1&<1&= ;01;0;
PAGE 192
R R R ',0(16,21 7 ,7/(f} &207f&07f & &20021 $;6 $9f<6$9f;01f ;0;f=01f <0;f';f'<^f'=^f} *59/ f '&1&f;$9f <$9f=$9f7.f=0;f88% f & &20021 $+6;0$;<0$;=0,1=0$;13,51'1&;1&<1&=67 2 &20021 $6,*86,*96, *:&&$;&&$<&&$=&&/;&&/<&&/= =2867$5%(7$3(/3767$575()$/3+$635()'3'=.735$=%$5 .8%58%5& & 135% 5($' ;6$9/f / 135%f<6$90f 0 135%f 5($' 7,7/(/f / Of&2070f 0 Of &07,1f 1 f 5($' 1367 5($' +6;0$;<0$;=0,1=0$;4 5($' 6,*86,*96, *:%;%<%=&$;&$<&$=&/0';&/0'< &/0'= & .8%5 )/$7 9(/ 352),/( & .735 &167$17 9(57,&$/ 0($1 $&&(/(5$7,21 5($' .8%5.735 ,).8%5(4Of 5($' 8%5& ,).735(4Of 5($' $=%$5 5($' =2(/386 7$5767$5$/3+$75()635()'3'= 5($' 1&;1&<1&= 5($' ;01 -f 1&;f^;0;.f} . 1&;f <0;/f / 1&;f=010f 0 1&;f=0;1f 1 1&;f 5($' ,51' & 35,17 7,7 / (/f / f &07Â0f 0 Of &071f 1 f 35,17,1* 7+( '$7$ 35,17 35,17 /;6$9/f<6$9/f / 135%f & 35,17 1367+6;0$;<0$;=0,1=0$;46,*86,*96,*: ,51'635()'3'= & &&$; &$;r%; &&$< &$
PAGE 193
$ ,).8%5(4f 35,17 O.8%58%5& , ) .7 35 fµ (fµ f 35,17 .7 35 $ = % $5 35,17 =2(/3867$5767$5$/3+$75()635()'3'= & 35,17 1&;1&< 1&= 35,17 /;01/f};0;/f<0;/f=01/f=0;/f / 1&;f & )250$7) f )250$7) f )250$7$$f )250$7,)f )250$7f )250$7$$$f $2 )250$7$;n352%$%,/,7< '$7$nn 3$57 12fµ;n1250$/n L n &ž)n;n(;317/ &')nf $ )250$7 ;, } ;} )fµ} ;)fµf )250$7$; r13 ; fµ67 fµ(;fµ+6 n(; n;0$; fµ(;fµ<0$; n(;m =0,1 r(;n=0$;n n n}(;n 4 n(m;r 6,*8 r(;r6,*9 r (;fµ6,*: n(; n ,51' n, n 635() fµ( $ ;n'3'= n(f )250$7$; fµ%; n )$; n%< fµ)$;n%= r) ;n&$; n); fµ&$< fµ) $; n&$= r)$ fµ &/0'; n)$$;&/0'< n)$$;n&/0'= n)$ $;f«7+( $%29( '$7$ *,9( 7+( )2//2:,1* 6,*0$ $1' /$0'$ fµ n9$/8(6n; n6,*$; n)n=n$; r 6,*$< n)fµ=m $ $;n6,*$= n),$n=n;n;/0'; n)$nr=r$;n;/0'< n ),$fµr=n;n;/0'= n)$nr=nf )250$7f¬ 1&; fµ;fµ1&< fµ, ;n1&= fµ, f )250$7n '2:1 :,1' ',67$1&( 180%(5 fµ,$fµ ;01 fµ(; n;0; n(;n<0; n(;n=01 r(;m=0; r (f )250$7;n.8%5 n,;fµ)/$7 9(/2&,7< 352) ,/(r;fµ8%5& fµ ( f )250$7;n.85 n n 0($1 $&& ,6 &2167$17 $=%$5 n (f )250$7; n= n(; fµ /35,0( n($;n867$5 n (O $; n76 7 $5 n($;m $/3+$ n($;fµ75() fµ (O$;n35() n(;n'3'= n(f 5(7851 (1'
PAGE 194
R R R R R Wrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrr`rrrAr`rArAr rrrrr +376 rrrrr 385326( 72 &$/&8/$7( 7+( *5,' ',0(16,216 $1' 7+( 1250$/,=,1* &2167$176 )25 &219(57,1* 7+( 3$57,&/( 6$03/,1*6 72 &21&(175$7,216 &rrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrr 68%5287,1( +376 &20021 $O;6$9Of<6$9 f ;01f;0;f=01f <0;f';f2
PAGE 195
R R R R R QRQR &rrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrr rrrrr &175 rrrrr 385326( 72 &$/&8/$7( &21&(175$7,216 )520 3$57,&/( 6$03/,1*6 rrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrr 68%5287,1( &175 ',0(16,21 180^f &20021 $O;6$9f<6$9f ;01f;0;f=01f <0;f';f'
PAGE 196
RRRRQQRR er 5:/. rrrrr 385326( 72 &$55< 287 5$1'20 :$/.6 )25 3$57,&/(6 86,1* $ 0217( &$5/2 0(7+2' *(1(5$7( 5$1'20 180%(56 :,7+ '(6,5(' 352%$%,/,7< ',675,%87,216 68%5287,1( 5:/. & &217,18286 6,08/$7,210$66 ,1-(&7,21 $7 +6 & *(1(5$/ 67$%,/,7< &21',7,216 & /$0'$7;/0' ! 9$5,(6 :,7+ +(,*+7 & $;,$/ ',))86,21 ,1&/8'(' & /$0'$7 ,1 7+( ; $1' < ',5(&7,216 $5( (48$/ & ,03/,&,7 ,17(*(5^3 f ',0(16,21 66$<$9f66$=$9}f66$;$9f & &20021 $O;6$9,f<6$9^f;01f;0;f=01,f <0;f';f'
PAGE 197
R R 6*$=0; &&$==22 & &2167$176 )25 /2*/,1($5 9(/2&,7< 352),/( 8& 867$5$. 8& %(7$3(/3r8& 8& 8&Or$/2*=2ff & &2167$176 )25 32:(5 /$: 9(/2&,7< 352),/( = $%6r(/3f ), $/2*=O=2-=2-=OA%(7$3(/3 ))) ),=Orr$/3+$ 8% 8&r))) & 35,17 $.*$00$%(7$3(1(1(17+538&8&8&= L ),)))8%, & ,1,7,$/,=$7,21 ;13 13 & 111 3$57,&/( 180%(5 111 1;1*; 1;1*< 1;1*= 1*51' & & 36,*1 326,7,9( $&&(/(5$7,21 3(5785%$7,21 & 36 ,*1 1(*$7,9( $&&(/(5$7,21 3(5785%$7,21 '2 66$;$9Â-f ;6$9-f 66$<$9,-f ;6$9-f 66$=$9,-f ;6$9-f 66$;$9,-f 66$;$9-f 66$<$9Â-f 66$<$9^ -f 66$=$9,-f 66$=$9-f & '2 -; 1&; '2 -< 1&< '2 -= 1&= &21&-;-<-=f & '2 ; 1&; '2 0 } '2 1 7.^01-;f & +$/) 7,0( 67(3 +67 67 32,17 2) 5(/($6( ; ;3 < = +6 =3 111 111
PAGE 198
R R 6 7 $8/ , 7$8/ /5*5 ,)111(4Of 35,17 ,)111(4 f 35,17 ,)111*713f *2 72 ,1,7,$/ 9(/2&,7< ,),.8%5(*f *2 72 ,)=/7=Of *2 72 8%$5 8%r=rr$/3+$ *2 72 8%$5 8&r$/*==f8&r=8& *2 72 8%$5 8%5& & ,51' ,51'r ,),51'*(2f *2 72 ,51' ,51'L 38 ,51'339 & ,51' ,51'r ,),51'*(2f *2 72 ,51' ,51' 39 ,51'339 & & ,51' ,51'r ,) Â,51'*(2f *2 72 ,51' ,51ž 3: ,51'339 83 ;6$938fr6,*8 8 8%$583 9 ;6$93 9fr6,*9 : ;6$93:fr6,*: & & ,1,7,$/ 7$8n6 & 7$8; ,6 $6680(' (48$/ 72 7$8< ;/0'< &&/
PAGE 199
,),51'*72f *2 72 ,51' ,51' 37$8= ,5L0'337 7$8= ;/0'=r<6$9^37$8=f 7$8< <6$937$8
PAGE 200
,51' ,51r ,),51'*(2f *2 72 ,51' ,51' 3$= ,51'33$ 6,*$= &&$== ,)=/7=22f 6,*$= 6*$=0; $=3 66$=$93$=36,*1=fr6,*$= & ,),.735(4f *2 72 = 3 =(/ 3 ,)=/7=Of *2 72 :21 )))r=rr$/3+$7+53 (1r=3 f *2 72 :*1 O$/2*==2f=2f%(7$3r=3f7+53O(1Or=3f 7:2 O(0Ofr=3^(1=3(1f $=%$5 *rO(1 r:217:2ff $= $=%$5$=3 ,)Â111(4$1';/7f 35,17 };<=:217:2$=%$5$=W $
PAGE 201
R Q R 8, 8 83, 83 & $9(5$*( 9(/2&,7< ,).8%5(4Of *2 72 ,)=/7=Of *2 72 8%$5 8%r=rr$/3+$ , 8%$5 8&r$/4* = =2f8&r=8& 83 83$;r)567 ;3 ;3r83,83fr)567 8 8%$58 3 ; ; 08W8fr)567 ,)^;/ (;0$;f *2 72 1;1*; 1;1*; *2 72 & 9 , 9 9 9$
PAGE 202
R R &217,18( ,)367$8(4167$8f *2 72 367 $8 367 $8 & 1(: 9(/2&,7< $1' 326,7,21 8, 8 83, 83 ,).8%5(*Of *2 72 ,)=/7=Of *2 72 8%$5 8%r=rr$/3+$ *2 72 8%$5 8&$/*==2f8&r=8& 83 83$;r67 ;3 ;3+67r83,83f 8 8%$583 ; ; +67r8,8 f ,);/(;0$;f *2 72 1;1*; 1;1*; *2 72 & 9 , 9 9 9$
PAGE 203
;3 ;3r83 ,83fr67$8/, 8 8%$583 ; ;r8,8fr67 $8/ , ,);/(;0$;f *2 72 1;1*; 1;1*; *2 72 & 9, 9 9 9$
PAGE 204
& 287387 35,17 131;1*; 1;1*< 1;1*= 1*51' & )250$7^fµ r }fµ 326,7,216 2) 7+( 3$57,&/(6 $7 7+( 7,0( 2)n n/($9,1* 7+( 5(*,21 2) ,17(5(67 ,1 < $12 = ',5(&7,216n ; }n111n ;n;n};fµ
PAGE 205
< ),77,1* *$866,$1 352),/( ,1 $67 648$5( ),76 :,7+ 7+( 67$&. &(175( 2) *5$9,7< (5(17, $7,1* 6,*0$6 ',67$1&(6 +( <',5(&7, 21 +( =',5(&7,21 & rrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrr & rrrrr 6,*0$6 rrrrr & 385326( & 72 &$/&8/$7( & f 6,*0$< $1' 6,*0$= % & 7+( =',5(&7, 21 %< /( & +(,*+7 $1' :,7+ 7+( & f ',))86,9,7,(6 %< ',)) & 120(1&/$785( & 1&; 12 2) '2:1 :,1' & 1&< 12 2) *5,'6 ,1 7 & 1&= 12 2) *5,'6 ,1 7 & 867$5 )5,&7,21 9(/2&,7< & 528*+1(66 /(1*7+ & 4 6285&( 675(1*7+ & + 67$&. +(,*+7 & ; '2:1 :,1' ',67$1& & =0,1 /2:(5 /,0,7 2) = & =0$; 833(5 /,0,7 2) = & '< *5,' 6,=( ,1 <', & '= *5,' 6,=( ,1 =', & 6& 680 2) &21&(175$7 & 680& 680 2) &21&(175$7 & =%$5 &(175( 2) *5$9,7< & ;; '2:1 :,1' ',67$1& ( *5,'6 $7 ; *5,'6 $7 ; 5(&7,21 5(&7,21 ,216 ,1 7+( ,216 ,1 7+( =',5(&7, 21 <',5(&7, 21 &rrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrr ',0(16,21 =f8^f680& 6,*
PAGE 206
R R R R R R & & &$/&8/$7,21 2) =%$5 66 M 66 '2 . 1&= 66 66=.fA$,<.f 6 6 6 6 $ , < . f & = %$5 6666 & 35,17 = %$5 &$/&8/$7,21 2) 6,*0$= :,7+ =%$5 , 7 &$// 3*6*0=4;8)1&==%$56*=$,<= ,7f 6*= =,3f 6*= 6(&21' 020(17 2) =352),/( &$// ==%641&==%$5==%564r$,<=f ==01 ,3f 6457==%564f ;;, ,3f [ 8&,3f 8) &217,18( ',))86,9,7,(6 35,17 & & <',5(&7,21 )520 6,*0$< 35,17 ¯ ;; ,f 1&;f &$// ',)671&;;;6,*<8&867$5+=2f & =', 5(&7, 21 )520 3* 6,*0$= 35,17 ;;,f 1&;f &$// ',)671&;;;6,*=8&867$5+=2f & =f§', 5(&7, 21 )520 6,*0$= :,7+ =%$5 35,17 ;; , f 1&;f &$// ',)671&;;;6*==8&867$5+=2f & =',5(&7, 21 )520 6(&21' 020(17 2) = ',675,%87,21 35,17 ;; ¯ ,f 1&;f &$// ',)671&;;;==018&}867$5+}=2f & )250$7fµr;n',))86,9,7,(6 2%7$,1(' %< ',))(5(17,$7,1*n n ',63(56,21 6&$/(6n;n; '2:1:,1' ',67$1&(n 2;n. ',0(16, 21$/ ',))86,9,7< )520 ',))(5(17,$7,21n ;r.6 '(',0(16,1$/,=(' ' , )) .8r;fµ ;n.6 '(',0 ' , )) .8r+r;r.6 '(',0fµ fµ ',)) .8r6,*0$n ;nrr 60227+(' 9$/8(6 86(' )25 '(5,9$7,9(6 $5( /,67('r fµ rrnf )250$7f )250$7)f )50$7;(;ff
PAGE 207
)250$7 ;}7+( 581 3$5$0(7(56 $5( rrrr;r867$5 fµ (/;= n(;n4 mI(;+ (f )250$7O2;fµ; fµ(;fµ=0,1 fµ(O n=0$; r ;n'< fµ(;fµ'= fµ(f )250$7n =%$5 fµ(` )250$7Â,2;n.< )520 ',))(5(17,$7,1* 3* 6 ,*0$< r fµ ; fµ ;(,f;(,;ff )250$7,;}n.= )520 ',))(5(17,$7,1* 6,*0$= :,7+ + , n ; n;(,f;(;ff )50$7,;n.= )520 ',))(5(17,$7,1* 6,*0$= :,7+ =%$5r n ; r ;(,f;(,;ff )50$7,;}n.= )520 ',))(5(17,$7,1* 6(&21' 020(17 2) =r r ; n;(,f;(,;ff 6723 (1'
PAGE 208
RR R RR RQ RRRRRRRRRQRR AAAAAAAAAAAAAAAAAAAAAArrrrr rrrrrrrrrrrrrrrrrrrrrrrr rrrrrr rrrrrrr rrrrr <=8$9 rrrrr 385326( &$/&8/$7,21 2) ,17(*5$/6 &2< $12 &'= 0$66 %$/$1&( $1' ',6&5(7,=$7,21 2) 9(/2&,7< 352),/( 120(1&/$785( 8) &219(&7,21 9(/2&,7< < &(175( 2) *5,' ,1 7+( <',5(&7,21 = &(175( 2) *5,& ,1 7+( =',5(&7,21 $ ,= ,17(*5$/ &&= $ , < r ,17(*5$/ &2< 8 ',6&5(7,=(' 9$/8(6 2) 9(/2&,7< e Y Rn r e r 68%5287,1( <=8$91&<1&=867$5=&4=0,1r'<&=8) 6&681&<$,==$,
PAGE 209
35,17 '2 . }1& = $,<.f %r680&.fr&< 35,17 =.f 8.f680&.f$,<.f )250$7^;63$&( $9(5$*(' 9$/8(6r;f«=n;n8};W I680&nWO;rI,
PAGE 210
R R R R R R Q &rrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrr]rrrrrrrrrrrr rrrrr 3*6*0< rrrrr 385326( &$/&8/$7,21 2) 6,*0$< 120(1&/$785( 6*< 6,*0$< (< *$866,$1 9$/8( 2) <)81&7,21 )< &$/&8/$7(& 9$/8( &) <)81&7,21 &rrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrArrrrrArrrrrrrrrrrr 68%5287,1( 3*6*0<4;8)1&<6*<$ ,=
PAGE 211
RRRRRRRQRR rrrrr 3*6*0= rrrrr 385326( &$/&8/$7,21 2) 6,*0$= 120(1&/$785( 6*= 6,*0$= (= *$866,$1 9$/8( 2) =)81&7,21 )= &$/&8/$7(& 9$/8( &) =)81&7,21 += '(9,$7,21 2) )= )5&0 (= ,7 ,) 6,*0$6 &$/&8/$7(& :,7+ 67$&. +(,*+7 ,7 ,) 6,*0$6 &$/&8/$7(& :,7+ &(175( &) *5$9,7< &rrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrr 68%5287,1( 3*6*0=4;8)1&=+6*=$ , <=,7f ',0(16,21 )=f(=fr+=*f}5&/f,1$f}=f$,
PAGE 212
R R R R R R + =.f ) =.f f§ ( =.f 35,17 }=.f})=.f(=.f}+=.f )250$7;}fµ3f§* 6,*0$= .,7( 8) $1' + n(, )250$7 &;} r 3* 6,*0$= K,7+ 8) $1' =%$5 fµ ( f )250$74;Wn=nWLJ;Wn)=nWO%;n(=nWO%;Wn+=nf )250$7;(;ff )250$7n rr (5525 rrr ,7(67 r,f 5(7851 (1' rrrrr ]]HV4 rrrf«r 385326( &$/&8/$7,21 2) 6(&21' 020(17 &) =' ,675,187,21 120(1&/$785( ==01 6(&21' 020(17 2) =',675 = =(564 ==01rr &rrrr rrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrr -rrrrrrrr 68%5287,1( ==%64,1&==%$5==(564$ , <=f ',0(16,21 =&f$,
PAGE 213
R R R R R R R &rrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrArrA rrrrr ',)67 rrrrr 385326( &$/&8/$7,21 2) $9(5$*( &,))86 ,9 ,7,(6 120(1&/$785( 8& &219(&7,21 9(/2&,7< $55$< 2) /(1*7+ 1&; *(7 ;&(5,9$7,9( 2) 6,*0$6 7. $9(5$*( ',))86,9,7< &rrrr rrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrr rrrrrreer 68%5287,1( ',)671&;;;6*8&867$5+=2f & ',0(16,21 ;;f6*f8&f7.^f'(7f & &$// 6*;;6*6*1&;,(5f , ) ,(51(2f *& 72 ,& 35,17 6*, ,f 1&;f & &$// &*7;;6*'(71&;,(5f ,) ,(51(2f *& 7& & 35,17 *(7 ,f 1&;f & '2 1&; 7 .^,f '( 7,fr8& , fr6* , ! 35,17 7. ,f 1&;f & '& & 1&; 7.,f 7.,-867$5;;,f 35,17 7. ,f 1&;f & '2 1&; 7.,f 7. , fr;;^,f+ 35,17 7. ,f 1&;f & '2 1&; 7.,f 7. , fr+6*, f 35,17 7. ,f 1&;f *2 7& & 35,17 ,(5 &$// 37')1&;867$5=+;;6**(7f & )250$7n )250$7n )250$7f¬ )250$7f¬ )250$7r )250$7n 6* &(5 . .6 .6 .6 ; ( ;( ,;( ;( ;( ;( f ; f ; f ; f ; f ; f; ( ^ ( ( ( ( ( ,;ff ,;f f ,;ff ,;f f ,;f f ,;ff )250$7fµrrr (5525 rrr ,(5 f¬,f 5(7851 (1'
PAGE 214
R R R R R &rrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrr rrrrr 372) rrrrr 385326( &$/&8/$7,21 2) 32,17 &,))86 ,9 ,7 , (6 120(1&/$785( 7.3 32,17 &,))86,9,7< 68%5287,1( 37')1&;867$5=2+;;6*'(7 f ',0(16,21 =f7.3f;;f&(7f6*f8f & 8 867$5&r$/*= f 8 867 $52 & '2 1&; & 867$5r;;, f & +;;^ , f & 6* , f+ 35,17 ;; , f =f & =^/f =fr & '2 . ,).1(Of =.f =.f= Of 8.f 8r$/**=.f=f8 7.3.f 8.fr6*,fr'(7,f 35,17 =. f } . f 8. f . f ^7.3.f . Of & '2 . 7.3.f 7.3.f&O 35,17 7.3.f . f & '& . 7.3^.f 7.3.f& 35,17 7.3.f . Of & '2 . 7.3.f 7.3.f& 35,17 7.3.f . OLf & ,)=f/7f *2 72 &217,18( )250$7; fµ32,17 & , ))86 ,9 , 7 , (6 3523&5$7 ,&1$/ 72 32,17r n 9(/2&,7,(6r;n$7 ; m(! )250$7n = n ; ( ffµ 8 n,;(OOf n . n ,;(OOff )250$7n .6 n ,;(OOff )250$7r .6 fµ ; ( f f )250$7n .6 fµ ,;(O f f 5(7851 (1'
PAGE 215
/,7(5$785( &,7(' $QJHO¯ -. 0HDVXUHPHQW RI /DJUDQJLDQ DQG (XOHULDQ SURSHUWLHV RI WXUEXOHQFH DW D KHLJKW RI IW 4XDUW -U 5R\ 0HW 6RF f $QJHO¯ -. 3DFN '+ +RHFNHU :+ DQG 'HOYHU 1 /DJUDQJLDQ (XOHULDQ WLPH VFDOH UDWLRV HVWLPDWHG IURP FRQVWDQW YROXPH EDOORRQ IOLJKWV SDVW D WDOO WRZHU 4XDUW -U 5R\ 0HW 6RF f $U\D 636 6WUXFWXUH RI VWDEO\ VWUDWLILHG WXUEXOHQW ERXQGDU\ OD\HU 86 $UP\ 5HVHDUFK *UDQW '$$0&* &ROOHJH RI (QJLQHHULQJ &RORUDGR 6WDWH 8QLYHUVLW\ )RUW &ROOLQV &ROn RUDGR f $]L] . &DUU /$ 5RZH 5' DQG :DOOLV &RPSXWHU PRGHOLQJ RI SROOXWLRQ IURP DWPRVSKHULF VRXUFHV SDSHU SUHVHQWHG DW WKH WK MRLQW &KHP (QJU FRQI 9DQFRXYHU &DQDGD f %DOGZLQ /9 DQG 0LFNHOVHQ :5 7XUEXOHQW GLIIXVLRQ DQG DQHPRn PHWHU PHDVXUHPHQWV $6&( 7UDQV 3DUW f %DUDG 0/ 3URMHFW 3UDLULH *UDVV D ILHOG SURJUDP LQ GLIIXVLRQ 9RO *HRSK\ 5HV 3DS 1R $)&5/ 5HSW 75 Lf f %DUDG 0/ 3URMHFW 3UDLULH *UDVV D ILHOG SURJUDP LQ GLIIXVLRQ 9RO *HRSK 5HV 3DS 1R $)&5/ 5HSW 75 LLf f %DUELQ $5 DQG -RQHV -% 7XUEXOHQW IORZ LQ WKH LQOHW UHJLRQ RI D VPRRWK SLSH -U %DVLF (QJU $60( 7UDQV 6HU ( f %DWFKHORU *. 'LIIXVLRQ LQ D ILHOG RI KRPRJHQHRXV WXUEXOHQFHf§, (XOHULDQ DQDO\VLV $XV -U 6F 5HV 6HU $ B f %DWFKHORU *. 'LIIXVLRQ LQ D ILHOG RI KRPRJHQHRXV WXUEXOHQFHf§,, 7KH UHODWLYH PRWLRQ RI SDUWLFOHV &DPE 3KLO 6RF 3URF f %DWFKHORU *. 'LIIXVLRQ LQ IUHH WXUEXOHQW VKHDU IORZV -U )LG 0HFK f %DWFKHORU *. 6RPH UHIOHFWLRQV RQ WKH WKHRUHWLFDO SUREOHPV UDLVHG DW WKH V\PSRVLXP $GY *HRSK B f
PAGE 216
%DWFKHORU *. 7KH WKHRU\ RI KRPRJHQHRXV WXUEXOHQFH &DPEULGJH 8QLYHUVLW\ 3UHVV 1HZ
PAGE 217
%XOOLQ -$ +\EULG FRPSXWHU VLPXODWLRQ RI WXUEXOHQW GLIIXVLRQ LQ WKH DWPRVSKHUH E\ 0RQWH &DUOR PHWKRGV 3K' WKHVLV 8QLn YHUVLW\ RI +RXVWRQ +RXVWRQ 7H[DV f %XOOLQ -$ DQG 'XNOHU $( 6WRFKDVWLF PRGHOLQJ RI WXUEXOHQW GLIn IXVLRQ ZLWK K\EULG FRPSXWHU (QY 6F t 7HFK A f &DOGHU ./ (GG\ GLIIXVLRQ DQG HYDSRUDWLRQ LQ IORZ RYHU DHRG\QDP LFDOO\ VPRRWK DQG URXJK VXUIDFHV 4XDUW -U 0HFK $SSO 0DWK f &DOGHU ./ 6RPH UHFHQW %ULWLVK ZRUN RQ WKH SUREOHP RI GLIIXVLRQ LQ WKH ORZHU DWPRVSKHUH LQ $LU 3ROOXWLRQ 3URF 86 7HFK &RQI $LU 3ROO 0F*UDZ +LOO 1HZ
PAGE 218
&RUUVLQ 6 0HP /RFDO LVRWURS\ LQ WXUEXOHQW VKHDU IORZ 1$&$ 5HV % f &RUUVLQ 6 7XUEXOHQW IORZ $P 6F B f &UDPHU +( DQG 5HFRUG )$ 3URSHUWLHV RI WXUEXOHQW HQHUJ\ VSHFn WUD DQG FRVSHFWUD LQ WKH DWPRVSKHULF VXUIDFH OD\HU 5 t ' 7HFK 5HSW (&20*) 86 $UP\ (OHFWURQLFV &RPPDQG $WPRV 6F /DE 5HV 'LY )RUW +XDFKXFD $UL]RQD f &UDPHU +( 5HFRUG )$ DQG 7LOOPDQ -( ,QYHVWLJDWLRQ RI ORZ OHYHO WXUEXOHQW VWUXFWXUH IRU YDULRXV VXUIDFH URXJKQHVVHV DQG JURVV PHWHRURORJLFDO FLUFXODWLRQV 5HSW XQGHU FRQWUDFW '$$52 86 $UP\ 5HV 2II (QY 6F 'LY $UOLQJn WRQ 9D f &UDPHU +( 5HFRUG )$ 7LOOPDQ -( DQG 9DXJKQ +& 6WXGLHV RI WKH VSHFWUD RI WKH YHUWLFDO IOX[HV RI PRPHQWXP KHDW DQG PRLVWXUH LQ WKH DWPRVSKHULF ERXQGDU\ OD\HU $QQXDO 5HSW XQGHU FRQWUDFW 1R '$6& 'HSDUWPHQW RI 0HWHRUn RORJ\ 0,7 &DPEULGJH 0DVV f &URNH ( &KLFDJR DLU SROOXWLRQ V\VWHP PRGHO WKUHH TXDUWHUO\ SURJUHVV UHSRUWV $UJRQQH 1DW /DE 5HSW $1/(6&& f &VDQDG\ *7 7XUEXOHQW GLIIXVLRQ LQ WKH HQYLURQPHQW ' 5HLGHO 3XE &R %RVWRQ 0DVV f 'DYLGVRQ % $ VXPPDU\ RI 1HZ
PAGE 219
'\HU $7KH IOX[ JUDGLHQW UHODWLRQ IRU WXUEXOHQW KHDW WUDQVIHU LQ WKH ORZHU DWPRVSKHUH 4XDUW -U 5R\ 0HW 6RF B f (OOLVRQ 7+ 7XUEXOHQW GLIIXVLRQ 6F 3URJ f (OOLVRQ 7+ 7KH XQLYHUVDO VPDOO VFDOH VSHFWUXP RI WXUEXOHQFH DW KLJK 5H\QROGV QXPEHUV 0HFKDQLTXH GH OD 7XUEXOHQFH (GLWLRQV GX &HQWUH 1DWLRQDO GH OD 5HFKHUFKH 6FLHQWLILTXH 3DULV f (VFKHQURHGHU $4 DQG 0DUWLQH] -5 $QDO\VLV RI /RV $QJHOHV DWn PRVSKHULF UHDFWLRQ GDWD IURP DQG *HQ 5HV &RUS &5 f )DEULFN $0DWKHPDWLFDO PRGHOLQJ RI UHDFWLYH SRLQW VRXUFHV SDSHU SUHVHQWHG DW WKH WK DQQ PHHW RI $3&$ 'HQYHU &RO f )DQ /7 DQG +RULH < 5HYLHZ RI DWPRVSKHULF GLVSHUVLRQ DQG XUEDQ DLU SROOXWLRQ PRGHOV &5& &ULWLFDO 5HYLHZV LQ (QYLURQPHQWDO &RQWURO f )DYUH $ *DYLJOLR DQG 'XPDV 5 6RPH PHDVXUHPHQWV RI WLPH DQG VSDFH FRUUHODWLRQ LQ ZLQG WXQQHO 1$&$ 7HFK 0HP f )DYUH $ *DYLJOLR DQG 'XPDV 5 6SDFHWLPH GRXEOH FRUUHODWLRQV DQG VSHFWUD LQ D WXUEXOHQW ERXQGDU\ OD\HU -U )LG 0HFK B f )DYUH $ *DYLJOLR DQG 'XPDV 5 )XUWKHU VSDFHWLPH FRUUHODWLRQV RI YHORFLW\ LQ D WXUEXOHQW ERXQGDU\ OD\HU -U )LG 0HFK B f )HOOHU : $Q LQWURGXFWLRQ WR SUREDELOLW\ WKHRU\ DQG LWV DSSOLFDWLRQV 9RO -RKQ :LOH\ t 6RQV 1HZ
PAGE 220
*LHORZ 5 6WRFKDVWLF PRGHOLQJ RI WXUEXOHQW KHDW DQG PDVV WUDQVn SRUW DQLVRWURSLF VKHDU IORZ 3K' WKHVLV 8QLYHUVLW\ RI )ORULGD *DLQHVYLOOH f *LIIRUG ) -U $ VLPXOWDQHRXV /DJUDQJLDQ(XOHULDQ WXUEXOHQFH H[n SHULPHQW 0RQWK :HDWK 5HY c f *LIIRUG ) -U 7KH UHODWLRQ EHWZHHQ VSDFH DQG WLPH FRUUHODWLRQV LQ WKH DWPRVSKHUH -U 0HW B f *LIIRUG ) -U 5HODWLYH DWPRVSKHULF GLIIXVLRQ RI VPRNH SXIIV -U 0HW f *LIIRUG ) -U 6PRNH SOXPHV DV TXDQWLWDWLYH DLU SROOXWLRQ LQGLFHV ,QW -U $LU 3ROO f *LIIRUG ) -U 8VH RI URXWLQH PHWHRURORJLFDO REVHUYDWLRQV IRU HVWLPDWLQJ DWPRVSKHULF GLVSHUVLRQ 1XFOHDU 6DIHW\ B f *ROGVWHLQ 6 2Q GLIIXVLRQ E\ GLVFRQWLQXRXV PRYHPHQWV DQG RQ WKH WHOHJUDSK HTXDWLRQ 4XDUW -U 0HFK $SSO 0DWK f *UDVV $6WUXFWXUDO IHDWXUHV RI WXUEXOHQW IORZ RYHU VPRRWK DQG URXJK ERXQGDULHV -U )LG 0HFK f *XSWD +& 'LIIXVLRQ E\ FRQWLQXRXV PRYHPHQWV -U 0DWK DQG 3K\V f *XUYLFK $6 DQG =XENRYVNLL 6/ (YDOXDWLRQ RI VWUXFWXUH FKDUDFn WHULVWLFV RI WHPSHUDWXUH IOXFWXDWLRQV LQ WKH DWPRVSKHUH ,]Y $WPRV DQG 2FHDQ 3K\V B f +DQQD 65 $ PHWKRG RI HVWLPDWLQJ YHUWLFDO HGG\ WUDQVSRUW LQ WKH SODQHWDU\ ERXQGDU\ OD\HU XVLQJ FKDUDFWHULVWLFV RI WKH YHUWLn FDO YHORFLW\ VSHFWUXP -U $WPRV 6F f +DQUDWW\ 7+HDW WUDQVIHU WKURXJK D KRPRJHQHRXV LVRWURSLF WXUn EXOHQW ILHOG -U $,&K( f +DQUDWW\ 7DQG )OLQW '/ 9HORFLW\ SURILOH IRU IXOO\ GHYHORSHG WXUEXOHQW IORZ LQ D SLSH -U $,&K( B f +DXJHQ '$ 3URMHFW 3UDLULH *UDVV D ILHOG SURJUDP LQ GLIIXVLRQ 9RO *HRSK 5HV 3DS 1R $)&5/ 5HSW 75 LLLf f +DXJHQ '$ .DLPDO -& DQG %UDGOH\ () $Q H[SHULPHQWDO VWXG\ RI 5H\QROGV VWUHVV DQG KHDW IOX[ LQ WKH DWPRVSKHULF OD\HU 4XDUW -U 5R\ 0HW 6RF B f
PAGE 221
+D\ -6 DQG 3DVTXLOO ) 'LIIXVLRQ IURP D IL[HG VRXUFH DW D KHLJKW RI D IHZ KXQGUHG IHHW LQ WKH DWPRVSKHUH -U )LG 0HFK f +D\ -6 DQG 3DVTXLOO ) 'LIIXVLRQ IURP D FRQWLQXRXV VRXUFH LQ UHODWLRQ WR WKH VSHFWUXP DQG VFDOH RI WXUEXOHQFH $GY *HRSK f +HLQHV 76 DQG 3HWHUV /. 7KH HIIHFW RI D KRUL]RQWDO LPSHUYLRXV OD\HU FDXVHG E\ D WHPSHUDWXUH DORIW RQ WKH GLVSHUVLRQ RI SROn OXWDQWV LQ WKH DWPRVSKHUH $WPRV (QY @B Df +HLQHV 76 DQG 3HWHUV /. $Q DQDO\WLFDO LQYHVWLJDWLRQ RI WKH HIIHFW RI D ILUVW RUGHU FKHPLFDO UHDFWLRQ RQ WKH GLVSHUVLRQ RI SROOXWDQWV LQ WKH DWPRVSKHUH $WPRV (QY B Ef +HLQVHQEHUJ : =XU VWDWLVFKHQ WKHRULH GHU WXUEXOHQ] = 3K\VLN f +HVV 6/ ,QWURGXFWLRQ WR WKHRUHWLFDO PHWHRURORJ\ +ROW (LQHKDUW DQG :LQVWRQ 1HZ
PAGE 222
-RKQVRQ :% /XGZLJ )/ 'DEEHUGW )/ DQG $OOHQ 5$Q XUEDQ PRGHO IRU FDUERQ PRQR[LGH SDSHU SUHVHQWHG DW WKH WK DQQ PHHW RI $,&K( f .DKQ + $SSOLFDWLRQV RI 0RQWH &DUOR 86$(& 5HSW 505$1'f f .DLPDO $Q DQDO\VLV RI VRQLF DQHPRPHWHU PHDVXUHPHQWV IURP WKH &HGDU +LOO WRZHU $)&5/ (QY 5HV 3DS 1R f .HOORJ :: 'LIIXVLRQ RI VPRNH LQ WKH VWUDWRVSKHUH 0HW f .LUPVH ': *HQHUDOL]DWLRQ RI VWRFKDVWLF PRGHOV GHVFULELQJ WXUEXn OHQW PRWLRQ 06 WKHVLV ,RZD 6WDWH 8QLYHUVLW\ RI 6FLHQFH DQG 7HFKQRORJ\ $PHV ,RZD f .LUPVH ': $ 0RQWH &DUOR VWXG\ RI WXUEXOHQW GLIIXVLRQ 3K' WKHVLV ,RZD 6WDWH 8QLYHUVLW\ RI 6FLHQFH DQG 7HFKQRORJ\ $PHV ,RZD f .LUPVH ': DQG )DKLHQ 5: 0RQWH &DUOR SUHGLFWLRQ RI HGG\ GLIIX VLYLW\ WHQVRU FRPSRQHQWV SDSHU SUHVHQWHG DW 6DOW /DNH PHHWn LQJ RI $,&K( f .OHEDQRII 36 &KDUDFWHULVWLFV RI WXUEXOHQFH LQ D ERXQGDU\ OD\HU ZLWK ]HUR SUHVVXUH JUDGLHQW 1$&$ 7HFK 5HSW f .OHEDQRII 36 DQG 'LHKO =: 6RPH IHDWXUHV RI DUWLILFLDOO\ WKLFNn HQHG WXUEXOHQW ERXQGDU\ OD\HUV ZLWK ]HUR SUHVVXUH JUDGLHQW 1$&$ 7HFK 5HSW f .OXJ : 'LIIXVLRQ LQ WKH DWPRVSKHULF OD\HU FRPSDULVRQ RI VLPLn ODULW\ WKHRU\ ZLWK REVHUYDWLRQV 4XDUW -U 5R\ 0HW 6RF f .QLJKWLQJ ( 2Q WKH HTXDWLRQ RI GLIIXVLRQ LQ WKH DWPRVSKHUH 4XDUW -U 0HFK $SSO 0DWK B f .ROPRJRURY $1 $ UHILQHPHQW RI SUHYLRXV K\SRWKHVLV FRQFHUQLQJ WKH ORFDO VWUXFWXUH RI WXUEXOHQFH LQ YLVFRXV LQFRPSUHVVLEOH IOXLG DW KLJK 5H\QROGV QXPEHUV -U )LG 0HFK f .RRJOHU -% 6KROWHV 56 'DYLV $/ DQG +DUGLQJ &, $ PXOWL YDULDEOH PRGHO IRU DWPRVSKHULF GLVSHUVLRQ SUHGLFWLRQ -U $3&$ -/ f .RYDV]QD\ /6* 6SHFWUXP RI ORFDOO\ LVRWURSLF WXUEXOHQFH -U $HUR 6F f
PAGE 223
.UDLFKQDQ 5+ 7KH VWUXFWXUH RI LVRWURSLF WXUEXOHQFH DW YHU\ KLJK 5H\QROGV QXPEHUV -U )LG 0HFK f .UDLFKQDQ 5+ 'LIIXVLRQ E\ D UDQGRP YHORFLW\ ILHOG 3K\V )LGV f /DPE 5* 1RWH RQ WKH DSSOLFDWLRQ RI .WKHRU\ WR GLIIXVLRQ SUREOHPV LQYROYLQJ QRQ OLQHDU FKHPLFDO UHDFWLRQV $WPRV (QY B f /DPE 5* DQG 6HLQIHOG -+ 0DWKHPDWLFDO PRGHOLQJ RI XUEDQ SROOXn WLRQ (QY 6F t 7HFK B f /DQW] 5% &RDWV .0 DQG .ORHSIHU &9 $ WKUHH GLPHQVLRQDO QXPHULFDO PRGHO IRU FDOFXODWLQJ WKH VSUHDG DQG GLOXWLRQ RI DLU SROOXWDQWV SDSHU SUHVHQWHG DW WKH V\PS RQ DLU SROO WXUE DQG GLIIXVLRQ 1HZ 0H[LFR f ,QYHVWLJDWLRQ RI WXUEXOHQW IORZ LQ D WZRGLPHQVLRQDO FKDQ QHO 1$&$ 7HFK 5HSW f 7KH VWUXFWXUH RI WXUEXOHQFH LQ IXOO\ GHYHORSHG SLSH IORZ 1$&$ 7HFK 5HSW f + $ UHH[DPLQDWLRQ RI WKH n/HLS]LJ ZLQG SURILOHn FRQVLGHUn LQJ VRPH UHODWLRQV EHZWHHQ ZLQG DQG WXUEXOHQFH LQ WKH IULFn WLRQDO OD\HU 7HOOXV A f ++ :LQG SURILOH VXUIDFH VWUHVV DQG JHRVWURSKLF GUDJ FRn HIILFLHQWV LQ WKH DWPRVSKHUH VXUIDFH OD\HU $GY *HRSK A f /LHSPDQQ +: /DXIHU DQG /LHSPDQ . 2Q WKH VSHFWUXP RI LVRn WURSLF WXUEXOHQFH 1$&$ 7HFK 1RWH f /LHSPDQQ +: DQG 5RELQVRQ 06 &RXQWLQJ PHWKRGV DQG HTXLSPHQW IRU PHDQ YDOXH PHDVXUHPHQWV LQ WXUEXOHQFH UHVHDUFK 1$&$ 7HFK 1RWH f /OR\G $ 7KH JHQHUDWLRQ RI VKHDU IORZ LQ D ZLQG WXQQHO 4XDUW -U 5R\ 0HW 6RF B f /XFDV '+ 7KH DWPRVSKHULF SROOXWLRQ RI FLWLHV ,QW -U $LU 3ROO , f /XPOH\ -/ 7KH VSHFWUXP RI QHDUO\ LQHUWLDO WXUEXOHQFH LQ D VWDEO\ VWUDWLILHG IOXLG -U $WPRV 6F f /XPOH\ -/ DQG 3DQRIVN\ +$ 7KH VWUXFWXUH RI DWPRVSKHULF WXUEXn OHQFH ,QWHUVFLHQFH 1HZ
PAGE 224
0LOOHU 0 DQG +RO]ZRUWK * $Q DWPRVSKHULF GLIIXVLRQ PRGHO IRU PHWURSROLWDQ DUHDV -U $3&$ f 0RQLQ $6 6PRNH SURSDJDWLRQ LQ WKH VXUIDFH OD\HU RI WKH DWPRVn SKHUH $GY *HRSK M f 0RQLQ $6 DQG 2EXNKRY $0 %DVLF ODZV RI WXUEXOHQW PL[LQJ LQ WKH JURXQG OD\HU RI WKH DWPRVSKHUH $NDGHPLD 1DXN 6665 /HQLQJUDG f 0RQLQ $6 DQG
PAGE 225
3DQRIVN\ +$ DQG 0F&RUPLFN 5$ 7KH VSHFWUXP RI YHUWLFDO YHORFLW\ QHDU WKH VXUIDFH 4XDUW -U 5R\ 0HW 6RF A f 3DQRIVN\ +$ DQG 3UDVDG % 7KH HIIHFW RI PHWHRURORJLFDO IDFWRUV RQ DLU SROOXWLRQ LQ D QDUURZ YDOOH\ -U $SSO 0HW B f 3DU]HQ ( 6WRFKDVWLF SURFHVVHV +ROGHQ'D\ 6DQ )UDQFLVFR &DOLI f 3DVTXLOO ) 7KH HVWLPDWH RI WKH GLVSHUVLRQ RI ZLQGERUQH PDWHULDO 0HW 0DJ f 3DVTXLOO ) $WPRVSKHULF GLIIXVLRQ ' 9DQ 1RVWUDQG /RQGRQ f 3DWDQNDU 69 DQG 6SDOGLQJ '0 +HDW DQG PDVV WUDQVIHU LQ ERXQGDU\ OD\HUV &KHP 5XEEHU &R &OHYHODQG 2KLR f 3KLOOLSV 20 6KHDU IORZ WXUEXOHQFH $QQ 5HY )LG 0HFK M/ f 3RROHU ) $ SUHGLFWLRQ PRGHO RI PHDQ XUEDQ SROOXWLRQ IRU XVH ZLWK VWDQGDUG ZLQG URVHV ,QW -U $LU :DW 3ROO B f 3UDQGWO / ,QYHVWLJDWLRQV RQ WXUEXOHQW IORZ =DPP B f 3ULHVWOH\ &+% 7XUEXOHQW WUDQVIHU LQ WKH ORZHU DWPRVSKHUH 8QLY RI &KLFDJR 3UHVV &KLFDJR ,OO f 3ULHVWOH\ &+% 7HPSHUDWXUH IOXFWXDWLRQV LQ WKH DWPRVSKHULF ERXQGn DU\ OD\HU -U )LG 0HFK B f 3URXGPDQ , DQG 5HLG :+ 'HFD\ RI D QRUPDOO\ GLVWULEXWHG DQG KRPRn JHQHRXV WXUEXOHQW YHORFLW\ ILHOG 3KLO 7UDQV $ f 3UXSSDFKHU +5 DQG 1HLOEXUJHU 0 'HVLJQ DQG SHUIRUPDQFH RI WKH 8&/$ FORXG WXQQHO 3URF ,QW &RQI &ORXG 3K\V f 5HFRUG )$ DQG &UDPHU +( 7XUEXOHQW HQHUJ\ GLVVLSDWLRQ UDWHV DQG H[FKDQJH SURFHVVHV DERYH D QRQKRPRJHQHRXV VXUIDFH 4XDUW -U 5R\ 0HW 6RF B f 5H\QROGV $Q H[SHULPHQWDO LQYHVWLJDWLRQ RI WKH FLUFXPVWDQFHV ZKLFK GHWHUPLQH ZKHWKHU WKH PRWLRQ RI ZDWHU VKDOO EH GLUHFW RU VLQXRXV DQG RI WKH ODZ RI UHVLVWDQFH LQ SDUDOOHO FKDQQHOVf° 3KLO 7UDQV $ f
PAGE 226
5H\QROGV :& 5HFHQW DGYDQFHV LQ WKH FRPSXWDWLRQ RI WXUEXOHQW IORZV LQ $GYDQFHV LQ &KHPLFDO (QJLQHHULQJ 'UHZ 7% &RNHOHW *5 +RRSHV -: -U DQG 9HUQHXOHQ 7 (GV $FDGHPLF 3UHVV 1HZ
PAGE 227
6DQGERUQ 9$ DQG %UDXQ :+ 7XUEXOHQW VKHDU VSHFWUD DQG ORFDO ,VRWURS\ LQ WKH ORZVSHHG ERXQGDU\ OD\HU 1$&$ 7HFK 1RWH f 6DQGERUQ 9$ DQG 6ORJDQ /RQJLWXGLQDO WXUEXOHQW VSHFWUXP VXUn YH\ RI ERXQGDU\ OD\HUV LQ DGYHUVH SUHVVXUH JUDGLHQWV 1$&$ 7HFK 1RWH f 6FKOLFKWLQJ + %RXQGDU\ OD\HU WKHRU\ 0F*UDZ +LOO 1HZ
PAGE 228
6Q\GHU :+ DQG /XPOH\ -/ 6RPH PHDVXUHPHQWV RI SDUWLFOH YHORFLW\ DXWRFRUUHODWLRQ IXQFWLRQV LQ D WXUEXOHQW IORZ -U )LG 0HFK f 6WHUQ $& (G 3URFHHGLQJV RI V\PSRVLXP RQ PXOWLSOHVRXUFH XUEDQ GLIIXVLRQ PRGHOV $3&2 3XE 1R $3 f 6WHZDUW 1* *DOH +DQG &URRNV 51 7KH DWPRVSKHULF GLIIXVLRQ RI JDVHV GLVFKDUJHG IURP WKH FKLPQH\ RI WKH +DUZHOO UHDFWRU %(32 ,QW -U $LU 3ROO -/ f 6WUDWRQRYLFK 5/ &RQGLWLRQDO 0DUNRY SURFHVVHV DQG WKHLU DSSOLFDWLRQ WR WKH WKHRU\ RI RSWLPDO FRQWURO $PHULFDQ (OVHYLHU 1HZ
PAGE 229
7D\ORU 7D\ORU 7D\ORU 7D\ORU *, 'LIIXVLRQ E\ FRQWLQXRXV PRYHPHQWV 3URF /RQ 0DWK 6RF f *, 6WDWLVWLFDO WKHRU\ RI WXUEXOHQFH, 3URF 5R\ 6RF $ f *, 6WDWLVWLFDO WKHRU\ RI WXUEXOHQFH9 3URF 5R\ 6RF $ f *, 3URGXFWLRQ DQG GLVVLSDWLRQ RI YRUWLFLW\ LQ D WXUEXOHQW IOXLG 3URF 5R\ 6RF $ Df 7D\ORU *, 7KH VSHFWUXP RI WXUEXOHQFH 3URF 5R\ 6RF $ Ef 7D\ORU 57KH GLVVLSDWLRQ RI NLQHWLF HQHUJ\ LQ WKH ORZHVW OD\HUV RI WKH DWPRVSKHUH 4XDUW -U 5R\ 0HW 6RF BA f 7D\ORU 56LPLODULW\ WKHRU\ LQ WKH UHODWLRQ EHWZHHQ IOX[HV DQG JUDGLHQWV LQ WKH ORZHU DWPRVSKHUH 4XDUW -U 5R\ 0HW 6RF f 7HQQHNHV + DQG /XPOH\ -/ $ ILUVW FRXUVH LQ WXUEXOHQFH 7KH 0,7 3UHVV &DPEULGJH 0DVV f 7KRPDV ): &DUSHQWHU 6% DQG *DUWUHOO 6% 6WDFNVf§KRZ KLJK" -U $3&$ f 7KRPSVRQ 1 ,QWHQVLWLHV DQG VSHFWUD RI YHUWLFDO ZLQG IOXFWXDWLRQV DW KHLJKWV EHWZHHQ DQG IW LQ QHXWUDO DQG XQVWDEOH FRQGLWLRQV 4XDUW -U 5R\ 0HW 6RF B f 7RZQVHQG $$ 7KH PHDVXUHPHQW RI GRXEOH DQG WULSOH FRUUHODWLRQ GHn ULYDWLYHV LQ LVRWURSLF WXUEXOHQFH 3URF &DPE 3KLO 6RF f 7RZQVHQG $$ 7KH VWUXFWXUH RI WXUEXOHQW VKHDU IORZ &DPEULGJH 8QLYHUVLW\ 3UHVV &DPEULGJH (QJODQG f 7XFNHU +* $ JUDGXDWH FRXUVH LQ SUREDELOLW\ $FDGHPLF 3UHVV 1HZ
PAGE 230
9LQQLFKHQNR 1. 3LQXV 1= 6FKPHWHU 60 DQG 6KXU *1 7XUEXOHQFH LQ WKH IUHH DWPRVSKHUH &RQVXOWDQWV %XUHDX 1HZ
PAGE 231
%,2*5$3+,&$/ 6.(7&+ 6XUHVK % 3DKZD VRQ RI %DGUL 1DWK DQG 9LG\D 9DWL 3DKZD ZDV ERUQ RQ 1RYHPEHU DW 8MMDLQ ,QGLD ,Q $SULO KH ZDV JUDGXDWHG IURP 3DWXFN 7HFKQLFDO +LJK 6FKRRO %RPED\ ,QGLD +H HQUROOHG DW .& &ROOHJH %RPED\ 8QLYHUVLW\ IRU RQH DFDGHPLF \HDU EHIRUH WUDQVIHUULQJ WR ,QGLDQ ,QVWLWXWH RI 7HFKQRORJ\ %RPED\ LQ -XO\ ZKHUH KH UHFHLYHG D %DFKHORU RI 7HFKQRORJ\ GHJUHH LQ &KHPLFDO (QJLQHHULQJ LQ 0D\ ,Q 6HSWHPEHU KH HQUROOHG DW 8QLYHUVLW\ RI )ORULGD WR SXUVXH JUDGXDWH HGXFDWLRQ LQ WKH 'HSDUWn PHQW RI &KHPLFDO (QJLQHHULQJ +H ZDV PDUULHG LQ $XJXVW WR WKH IRUPHU %ULJLWWH 0DULH 5R¾DQ 8SRQ FRPSOHWLRQ RI KLV ZRUN IRU WKH GHJUHH RI 'RFWRU RI 3KLORVRSK\ KH ZLOO MRLQ ,17(5$ (QYLURQn PHQWDO &RQVXOWDQWV /WG +RXVWRQ 7H[DV
PAGE 232
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n(QJLQHHULQJ , FHUWLI\ WKDW , KDYH UHDG WKLV VWXG\ DQG WKDW LQ P\ RSLQLRQ LW FRQIRUPV WR DFFHSWDEOH VWDQGDUGV RI VFKRODUO\ SUHVHQWDWLRQ DQG LV IXOO\ DGHTXDWH LQ VFRSH DQG TXDOLW\ DV D GLVVHUWDWLRQ IRU WKH GHJUHH RI 'RFWRU RI 3KLORVRSK\ . ( *XEELQV 3URIHVVRU RI &KHPLFDO (QJLQHHULQJ
PAGE 233
, FHUWLI\ WKDW , KDYH UHDG WKLV VWXG\ DQG WKDW LQ P\ RSLQLRQ LW FRQIRUPV WR DFFHSWDEOH VWDQGDUGV RI VFKRODUO\ SUHVHQWDWLRQ DQG LV IXOO\ DGHTXDWH LQ VFRSH DQG TXDOLW\ DV D GLVVHUWDWLRQ IRU WKH GHJUHH RI 'RFWRU RI 3KLORVRSK\ WW ‘ 1$n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f§ F /eF 'HDQ &ROOHJH RI (QJLQHHULQJ 'HDQ *UDGXDWH 6FKRRO
PAGE 234
1RQ*DXVVLDQ H[SRQHQWLDO IXQFWLRQ )ROORZLQJ &HUPDNnV f GHYHORSPHQWV RI WKH VLPLODULW\ WKHRU\ IRU D JURXQG OHYHO VRXUFH &VDQDG\ f GHILQHV D OHQJWK VFDOH IRU WKH SOXPH VSUHDG DW ODUJH GLIIXVLRQ WLPHV DV A A GW 8) f :KHUH LV D FRQVWDQW &VDQDG\ DSSUR[LPDWHG E\ D SRZHU ODZ IXQFn WLRQ RI WKH GRZQZLQG GLVWDQFH DV < / FRQVW [ f F $VVXPLQJ WKDW WKH PDVV HGG\ GLIIXVLYLW\ KDV WKH VDPH =/Ff GHSHQGHQFH DV WKH HGG\ YLVFRVLW\ DQG DVVXPLQJ WKDW WKH VKHDU VWUHVV UHPDLQV FRQVWDQW LQ WKH JURXQG OD\HU DQG WKDW WKH YHORFLW\ SURILOH LV RI WKH SRZHU ODZ IRUP ZLWK WKH H[SRQHQW D KH GHULYHG WKH IROORZLQJ H[SUHVVLRQ IRU WKH YHUWLFDO HGG\ GLIIXVLYLW\ r ) . EB X / "f I f ] r F r U = ZKHUH
PAGE 235
DQG ) LV WKH FRQFHQWUDWLRQ GLVWULEXWLRQ LQ WKH YHUWLFDO GLUHFWLRQ ] DV D IXQFWLRQ RI DQG WKH SULPH GHQRWHV WKH GHULYDWLYH ZLWK UHn VSHFW WR $VVXPLQJ WKDW WKH VKHDU VWUHVV UHPDLQV FRQVWDQW LQ WKH JURXQG OD\HU DQG WKDW WKH PDVV HGG\ GLIIXVLYLW\ KDV WKH VDPH FW GHSHQGHQFH DV WKH HGG\ YLVFRVLW\ LH e KH REWDLQHG r ) ( FRQVW fµ e f§ f ) ] ZKLFK RQ LQWHJUDWLRQ JLYHV f ,Q JHQHUDO RQH PD\ H[SHFW WKH WXUEXOHQW 6FKPLGW QXPEHU HGG\ PDVV GLIIXVLYLW\HGG\ NLQHPDWLF YLVFRVLW\f WR EH D IXQFWLRQ RI WKH KHLJKW DQG WKXV PD\ EH ZULWWHQ DV B5 . FRQVW fµ IF ] f 7KHUHIRUH FRQVW f ZKLFK RQ LQWHJUDWLRQ JLYHV r ) ] ?& D% H[S f
PAGE 236
7KXV IRU HOHYDWHG VRXUFHV ZLWKRXW UHIOHFWLRQ RQH PLJKW ZULWH f ZKHUH D f 6XEVWLWXWLQJ (T f LQWR (T f LQWHJUDWLQJ LQ WKH \GLUHFWLRQ DQG WDNLQJ WKH QDWXUDO ORJDULWKP ZH REWDLQ OQ,\f ,Q FR Df ]+ / F f 7KH H[SRQHQW FDQ EH REWDLQHG E\ SORWWLQJ OQ>OQ,\f ,Q FT WW JAf @ fµN YV ,]+, DQG HYDOXDWLQJ WKH VORSH RI WKH FXUYH )RU ]T N N DQG [ DQG [ LH ZKHUH WKH UHIOHFWLRQ FDQ EH QHJOHFWHG LW LV VHHQ WKDW LV DFWXDOO\ D IXQFWLRQ RI WKH KHLJKW ] DQG RI WKH GRZQZLQG GLVWDQFH [ )LJV f DQG ff +RZHYHU DQ DYHUDJH YDOXH RYHU ] ! FDQ UHDVRQDEO\ GHVFULEH WKH GLVWULEXWLRQ 7KH YDOXH N DW [ REWDLQHG IRU WKLV FDVH LV ! 7KLV YDOXH FRPSDUHV ZLWK D YDOXH RI IRU D *DXVVLDQ GLVWULEXWLRQ DQG D YDOXH RI IURP VLPLODULW\ WKHRU\ ZLWK D 7KHQ IRU VRXUFHV ZLWK UHIOHFWLRQ
PAGE 237
81,9(56,7< RI 8) )/25,'$ 7KH )RXQGDWLRQ IRU 7KH *DWRU 1DWLRQ ,QWHUQHW 'LVWULEXWLRQ &RQVHQW $JUHHPHQW ,Q UHIHUHQFH WR WKH IROORZLQJ GLVVHUWDWLRQ $87+25 3DKZD 6XUHVK 7,7/( 6WDWLVWLFDO PRGHOLQJ DQG VLPXODWLRQ RI DWPRVSKHULF WXUEXOHQFH UHFRUG QXPEHU f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f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
|
|