Lake Bloomington TMDL
Nov 2006 Final Stage 1 – Characterization Report 6-3
6.3 MODEL CALIBRATION AND VALIDATION
Calibration involves minimizing the deviation between measured and simulated water quality indicators
by adjusting model parameters. Data required for calibration include a set of known input values along
with corresponding field observations. Although model calibration is critical, Tetra Tech believes that
significant effort should be focused on sound source characterization and sensitivity analysis. A good
characterization of source loadings results in a more efficient, scientifically sound, and justifiable
calibration process. Tetra Tech will identify data sets for water quality calibration, identify model
adjustment needs based on past experience, and work closely with IEPA to fully characterize sources and
address calibration issues and their impacts on final TMDL allocations. The performance of model
calibration will be assessed based on statistical methods and professional judgments.
Validation involves the use of a second set of independent information to check model calibration. Data
used for model validation consist of field measurements of the same type as the data output from the
model. Models are tested based on their predictions of mean values, variability, extreme values, and all
predicted values. If the model is calibrated properly, model predictions should be acceptably close to
field observations.
6.4 SENSITIVITY ANALYSIS
A thorough sensitivity analysis provides a number of benefits, including the following:
• Assistance on proper parameter selection
• Improved understanding of the model and related assumptions
• Evaluation of different TMDL scenarios
• Evaluation of model accuracy
• Justification of selection of Margin of Safety
The results of a sensitivity analysis will provide information regarding those parameters with the greatest
effect on outputs. Tetra Tech will perform a sensitivity analysis on multiple model runs based on selected
parameter range and load range. In addition to evaluating the sensitivity of the technical approach to the
different sources, it is also important to estimate (either qualitatively or quantitatively) the accuracy or
reliability of model predictions. This estimate of the model’s accuracy will be an important factor in
deciding how to use the model results in estimating the TMDL values.
An important step in the TMDL process is to evaluate the relative significance of the various source-loading
estimates on model results. For example, potential sources of total phosphorus contributing to the
impairment of the water body include municipal treatment plants, failing septic systems, livestock
operations, and urban runoff. It will be important to evaluate the sensitivity of the model to loadings from
each of these sources. For example, there is no known relationship that can be used to predict the
contribution of failing septic systems to a stream. If the analysis indicates that the model is especially
sensitive to this source, it might be necessary to revise the loading estimates to a daily or seasonal basis.
