Final Project

1
Final Project

• Summary of Results
• Conclusions

2
2008 Project TRUTH
Hydraulic Conductivity
Layer 1
PW2 discharge reduced To 0.90E8 ft3/year
3
Layer 2
4
Layer 3
All these complicated details may not
matter. What matters is to capture the essential
features of the system for the purposes of
predicting the response to pumping and the
movement of the contaminant particles.
5
Recharge rates
Extinction depths 10, 30 ft Leakance 4
ft/yr
6
PW2. Dry cell.
Reduce pumping rate from -0.99E8 ft3/year to
-0.90E8 ft3/year
7
PW1 doesnt capture any particles.
8
PW1
No cone of depression. PW1 doesnt look like a
sink and doesnt capture particles.
9
Particles by-pass PW1 and exit in PW2.
PW 2
All the particles exit in wells none end up in
the playa.
10
All of the particles that enter in layer 1, stay
in layer 1.
11
Dry cells. PW2 went dry.
Predicted ARM gt Calibrated ARM
12
Doesnt include PW2 since it is also a target.
Dry cells. PW2 went dry.

• 1. Predicted ARM gt Calibrated ARM
• Generally predicted ARM at non-target cells gt
  predicted ARM at target cells

13
724 Project Results
?
Includes results from 2006 and 4 other years
A good calibration does not guarantee an
accurate prediction.
14
Calibrated ARM of around 1.0 is a good
calibration.
2006 Project Results

• 1. Predicted ARM gt Calibrated ARM
• Predicted ARM at pumping wells gt predicted ARM
  at targets

Does not include PW2 since it is also a target.
15
Calibrated ARM of around 1.0 is a good
calibration.

• 1. Predicted ARM gt Calibrated ARM
• Predicted ARM at pumping wells gt predicted ARM
  at targets

Predicted ARM at targets gt predicted ARM at
pumping wells
16
2006 Project Results
Despite the relatively poor calibration, groups 4
and 5 managed to capture the essential features
of the system for the purpose of the prediction.
17
2008 Results
Doesnt include PW2 since it is also a target.
Dry cells. PW2 went dry.
18
2008 Particle Tracking
19
2008 Results
Particle Tracking Results travel time (yr) exit
location
?
?
?
?
PW2 went dry.
Low porosity gives high velocity which yields
short travel times.
20
2006 Project Results
Despite the relatively poor calibration, groups 4
and 5 managed to capture the essential features
of the system for the purpose of the prediction.
21
2006 Project Results
Particle Tracking Results travel time (yr) exit
location
?
?
?
?
6 hits
5 hits
22
2008 Results
Group 3 managed to capture the essential features
of the system for the best pumping prediction and
the best prediction of particle exit points, but
not travel times.
23
Observations
Generally predicted ARM at targets gt Calibrated
ARM
Generally, predicted ARM at pumping wells gt
Predicted ARM at nodes with targets
Head predictions are more robust (consistent
among different calibrated models) than transport
(particle tracking) predictions.
24
To use conventional inverse models/parameter
estimation models in calibration, you need to
have a pretty good idea of zonation (of K, for
example).
(New version of PEST with pilot points does not
need zonation as it works with continuous
distribution of parameter values.)
Also need to identify reasonable ranges for
the calibration parameters.
25
Calibration to Fluxes

• When recharge rate (R) is a calibration
  parameter, calibrating to fluxes can help in
  estimating K and/or R.

R was not a calibration parameter in our problem.
26
In this example, flux information helps calibrate
K.
q KI
K ?
H1
H2
27
or discharge information helps calibrate R.
R ?
28
In our example, total recharge is known/assumed
to be 7.14E08 ft3/year and discharge recharge.
All water discharges to the playa. Calibration to
ET merely fine tunes the discharge rates within
the playa area. Calibration to ET does not help
calibrate the heads and K values except in the
immediate vicinity of the playa.
29
Conclusions

• Calibrations are non-unique.

• A good calibration (even if ARM 0)
• does not ensure that the model will make
• good predictions.

• Field data are essential in constraining the
  model
• so that the model can capture the essential
• features of the system.

• Modelers need to maintain a healthy skepticism
• about their results.

• Need for an uncertainty analysis to accompany
• calibration results and predictions.