|
INTRODUCTION TO Visual
PEST-ASP
Visual PEST combines the powerful
parameter estimation capabilities of PEST-ASP with the
graphical processing and display features of WinPEST.
ABOUT PEST
PEST-ASP is the latest version of PEST, the
pioneer in model-independent parameter estimation. Since
it was first released over six years ago, PEST has
gained extensive use all over the world in many
different fields. Over this time it has undergone
continued development with the addition of many new
features that have improved its performance and utility
to a level that makes it uniquely applicable in just
about any modeling environment. PEST is now used
extensively for automated model calibration and data
interpretation in groundwater and surface water
hydrology, geophysics, geotechnical, mechanical and
mining engineering, as well as many other fields.
About WinPEST
WinPEST brings parameter estimation and
predictive analysis to life! Not only can you run PEST
in a friendly, intuitive Windows environment, you can
study a set of comprehensive, evolving run-time displays
of key PEST variables providing a wealth of information
on the progress of the parameter estimation process
being undertaken by PEST. When PEST has finished
execution, an additional suite of plots become available
displaying a full range of statistical and other
information on calibrated parameters, calibration
residuals and other PEST variables.
You will be amazed at what you can learn about your
model, the information content of your calibration
dataset, and the confidence you can place in model
results from an examination of WinPEST's high-impact
plots. You will truly wonder how you ever got along
without it!
Visual
PEST-ASP MAIN FEATURES
Here are just some of the features which make Visual
PEST truly unique.
Inversion Engine
PEST uses an extremely
robust numerical inverse-problem-solver with painstaking
attention to detail in all aspects of design and
implementation.
Parameter Bounds
The user is able to set
upper and lower bounds on parameters during the
calibration process, thus ensuring that estimated values
are in range.
User Intervention
PEST allows
user-intervention in the parameter estimation process
whereby troublesome parameters can be held for a while
and the most recent parameter upgrade repeated without
the need for extensive recalculation.
Parallel Processing
With Parallel PEST,
model runs can be distributed across a PC network.
Savings in overall optimization time can be enormous.
Utilities
PEST is accompanied by a suite
of powerful utilities which automate PEST setup and
carry out extensive error checking on all aspects of
PEST input dataset construction.
Data Handling
Observations and parameters
can be grouped for ease of property definition and
weights allocation, and to add flexibility to the
inversion process.
Predictive Analysis
PEST's unique
predictive analyzer allows the user to estimate the true
range of predictive nonuniqueness associated with a
calibrated model.
WinPEST
The WinPEST graphical user
interface displays an extensive range of run-time and
post-run data for easy analysis and postprocessing.
Experience
PEST has stood the test of
time. It has been used to carry out countless
calibration and data interpretation exercises in all
fields of science and engineering. Over its six-year
life it has undergone continuous improvement and
refinement to keep it at the cutting edge of
model-calibration technology.
Visual
PEST-ASP MODEL INDEPENDENCE
PEST brings parameter estimation technology to
all modelers by combining a powerful inversion
engine with the ability to communicate with a model
through the model's own input and output files. Thus
PEST can be used with any model without the need for
any programming. To calibrate a model, simply:
- inform PEST which numbers to adjust on model input
files,
- identify those numbers on model output files for
which there are corresponding field or laboratory
measurements, and
- instruct PEST how to run the model (assumed to be
command-line-driven).
PEST then takes control of the model, running it as
many times as it needs to while adjusting parameter
values until the discrepancies between model outputs and
corresponding field or laboratory measurements are as
small as possible in the weighted least squares sense.
It then lists optimal parameter values, an estimate of
the uncertainty associated with optimal parameter
values, best-fit model outcomes, model-to-measurement
residuals, and a suite of statistics related to the
optimal parameter and residual sets.
Visual PEST MODEL CALIBRATION & DATA
INTERPRETATION
With PEST you can turn any model into a powerful data
interpretation package. Model calibration is no longer
the time-consuming, frustrating and often fruitless
exercise that it used to be. The user is free to unleash
his/her creativity in the calibration and
data-interpretation process while PEST carries out the
numerically intensive calculations required to implement
his/her ideas. PEST allows a modeler to truly understand
the capacity that a dataset possesses for the estimation
of parameters governing the workings of a system, and
how supplementary data are most efficiently gathered in
order to increase that capacity.
The possibilities for creativity and elegance in data
interpretation and model calibration are truly enormous
with PEST. The "model" can be a batch file holding one
or many executables. Thus you can calibrate a model
using data gathered over noncontiguous time intervals;
you can undertake simultaneous calibration of a
steady-state and transient model, a flow and transport
model, multiple recharge models together with a flow
model, a flow model combined with regularization
functionality, and much, much more. Because there is no
limit to the number of model input files that PEST can
write and the number of model output files which PEST
can read, the possibilities for composite model
construction are limited only by a user's imagination.
Visual
PEST PREDICTIVE ANALYSIS
PEST-ASP introduces predictive analysis, the latest
development in parameter estimation technology. The
Predictive Analyzer is a revolutionary approach to
modeling that allows the modeler to actually calculate
the uncertainties in model predictions arising from
uncertainties in model parameters while ensuring the
model remains calibrated.
A common mistake in many modeling exercises is to
undertake "sensitivity analysis" after a model has been
calibrated in order to estimate the uncertainties in
model predictions. There are two problems with this
approach. The first is that when a parameter is varied
in order to test the effects of this variation on
predictive output, the model may become uncalibrated.
Thus the prediction cannot be considered a true model
prediction. The second problem is that the variation of
individual parameters by a small amount in order to
assess predictive uncertainty, may seriously
underestimate the extent to which parameters could
actually vary and still keep the model calibrated; the
trick is to vary not just one, but possibly many
correlated parameters together in such a way that the
variation of these parameters has virtually no effect on
model outcomes under calibration conditions. It is the
variation of this combination of parameters (rather than
each parameter individually) that must be undertaken to
perform true predictive analysis.
The unique PEST-ASP predictive analyzer allows the
modeler to vary parameters in such a way as to ensure
that the model remains calibrated while, at the same
time, testing the effect of this variation on key model
predictions. The modeler can ask PEST to calculate the
highest or lowest value of a key model outcome while at
the same time ensuring that the parameter values used to
make this prediction are such as to keep the model
calibrated. The repercussions for model deployment are
profound. Now the user can test best and worst case
scenarios with ease, design a fail-safe remediation
system and/or optimize the efficiency of a monitoring
network. Modeling will never be the same!
Predictive Analysis allows you to quantify the
uncertainties typically associated with modeling by
directly calculating the definitive uncertainty limits
on key model predictions.
Visual
PEST PARALLEL PROCESSING
PEST-ASP comes with sophisticated parallel processing
capabilities enabling it to distribute and manage model
runs across a network to significantly reduce
optimization times. Thus model calibration or predictive
analysis can now be undertaken using more parameters and
larger models than has hitherto been possible.
WinPEST
WinPEST's high impact and informative graphics allows
you to understand the calibration and predictive
analyses processes like never before. Through a series
of evolving run-time displays, you can tell at a glance
where the process is going and whether or not your
intervention may be required (see below). When PEST has
finished running, WinPEST presents a further array of
colorful and educational plots through which you can
examine parameter uncertainty and nonuniqueness, analyze
calibration residuals (either as a whole or in
user-defined groups), and much more besides.
WinPEST can be used interchangeable with PEST-ASP
(and previous versions of PEST) so you can import your
existing PEST datasets and give them a whole new lease
of life as they explode into color.
WinPEST USER INTERVENTION
Sometimes the model calibration or predictive
analysis process encounters numerical difficulties. If
these are hampering a PEST run, WinPEST's informative
displays not only make this plain but provide
information through which troublesome parameters
(normally insensitive and/or highly correlated
parameters) can be identified. With a few mouse clicks,
you can then halt PEST execution, hold the offending
parameters at their current values, and recalculate
improvements to the other parameters without having to
recompute the Jacobian matrix (the most time-consuming
aspect of the parameter estimation process). Using this
unique methodology you can, more often than not, get a
stalled calibration process "back on the rails" with
minimum wasting of computer time - often a big issue
with large and complex models.
| 
