A DRAINMOD-based method to estimate effects of drainage water management on annual nitrogen loss to surface water.

摘要

Effects of drainage water management (DWM) on nitrogen (N) losses to surface waters can be estimated by multiplying the change in subsurface drainage and surface runoff, as predicted by DRAINMOD, by long-term mean nitrogen (N) concentration, which is a function of local site conditions (climate, soil, cropping system, farming practices, and drainage system). Consistent with experimental observations from several sources, this approximate method assumes that the effect of DWM on N losses is proportional to its effect on drainage and surface runoff volumes. The reliability of the method was evaluated by comparing estimated annual N losses for two sites (one in North Carolina and one in Illinois) to those predicted by the process-based nitrogen model DRAINMOD-NII. The long-term N concentrations used in the approximate method were based on DRAINMOD-NII predictions. DRAINMOD-NII simulations predicted that DWM would reduce long-term average N losses to surface waters by 35% for continuous corn (CC) and 33% for a corn-wheat-soybean (CWS) rotation in North Carolina and by 31% and 26% for CC and a corn-soybean (CS) rotation in Illinois. The approximate method estimated the annual effect of DWM on N losses within 3 kg ha-1 of that predicted by DRAINMOD-NII in over 68% of the years for the CWS and CS rotations at both sites and in over 51% of the years for CC. Overestimated effects in some years were balanced by underestimated effects in others. However, relatively large errors occurred in about 25% of the years when N concentrations in the drainage water varied significantly from the long-term average. These errors typically occurred following periods when crop yields deviated significantly from the average. When these outliers were excluded from the analyses, the goodness of fit between annual N losses predicted by DRAINMOD-NII and annual losses estimated by the approximate method was substantially improved (Nash-Sutcliffe EF values ranging from 0.54 to 0.79). Although more research is needed to improve the approximate method when concentrations vary significantly from the long-term average, the results presented herein indicate that the method provides a reliable means of assessing impacts of DWM under normal conditions.