Probabilistic Model of Ammonia and Toxicity Status for Urban Lake

Abstract

The development of a probabilistic modeling framework to conduct a priori simulations (forecasts) of the concentration of total ammonia (T-NH3) and status with respect to toxicity standards and criteria is documented for ammonia-polluted Onondaga Lake, Syracuse, N.Y. This framework utilizes a previously tested mass balance nitrogen model, long-term (27 year) simulations, and the Monte Carlo technique. Variability is accommodated, based on long-term records, for several parameters and forcing conditions, including (1) tributary runoff, (2) in-lake vertical mixing, (3) in-lake rate of nitrification, (4) T-NH3 concentration in the effluent of a wastewater-treatment plant, (5) lake pH, and (6) lake temperature. The probabilistic framework performed well in simulating the broad variations in T-NH3 concentrations observed in the lake over a 6-year period (1989–1994). The major sources of the observed interannual variability in the lake's T-NH3 pool are demonstrated to be natural variations in runoff and the irregular occurrence of “nitrification events” during fall mixing. Predictions are presented in the form of frequency distributions of T-NH3 concentrations and violations/exceedances of the standard/criterion, consistent with the format of U.S. Environmental Protection Agency toxicity criteria. Application of the probabilistic model is recommended for the total maximum daily load analysis that will establish the T-NH3 limit for the wastewater-treatment plant effluent.