POLLUTANT LOAD ESTIMATES OF WATER QUALITY DATA USING GENETIC-ALGORITHM

摘要

Water quality data are collected less frequently than fl ow data because of the cost to collectand analysis, while water quality data corresponding to fl ow data is required to compute,to calibrate other hydrologic models, etc. Regression models are applicable to interpolatewater quality data corresponding to fl ow data. A regression model was suggested which iscapable to consider fl ow and time variance, the regression model coeffi cients were calibratedusing various measured water quality data with genetic-algorithm. Both LOADEST and theregression using genetic-algorithm were evaluated using 19 water quality datasets throughcalibration and validation. The regression model using genetic-algorithm displayed similarmodel behaviors to LOADEST. The load estimates by both LOADEST and the regressionmodel using genetic-algorithm indicated that more water quality data does not necessarilylead to the load estimates with smaller error compared to measured load. Regression modelsneed to be calibrated and validated before they are used to interpolate pollutant loads, asseparating water quality data into two data sets for calibration and validation.