LEAST-SQUARES CALIBRATION OF QUAL2E

In recent years, much research has been directed toward developing objective, statistically valid methods for water quality model calibration (parameter estimation). Few of these methods have found their way into routine practice. This paper describes the application of a nonlinear regression technique to calibrate the popular water quality model, QUAL2E. A nonlinear programming model was developed to minimize the sum of squares of differences in model predictions and pollutant observations. QUAL2E is called as a subroutine by the program. Six parameters are simultaneously estimated for each of two intensive survey data sets. Because optimal parameter estimates were found to be considerably different for each data set, suggesting that the parameters were not true constants, these estimates were used to develop probability distributions for the parameters for use in a Monte Carlo model.