A COMPARATIVE-STUDY OF LINEAR AND NONLINEAR TIME-SERIES MODELS FOR WATER-QUALITY

Surface water quality data are routinely collected in river basins by state or federal agencies. The observed quality of river water general ly reflects the overall quality of the ecosystem of the river basin. A dvanced statistical methods are often needed to extract valuable infor mation from the vast amount of data for developing management strategi es. Among the measured water quality constituents, total phosphorus is most often the limiting nutrient in freshwater aquatic systems. Relat ively low concentrations of phosphorus in surface waters may create eu trophication problems. Phosphorus is a non-conservative constituent. I ts time series generally exhibits nonlinear behavior. Linear models ar e shown to be inadequate. This paper presents a nonlinear state-depend ent model for the phosphorous data collected at DeSoto, Kansas. The no nlinear model gives significant reductions in error variance and forec asting error as compared to the best linear autoregressive model ident ified.