Impact of rainfall data source on hydrologic and water quality model response of a coastal plain watershed

Abstract

Hydrologic and water quality models are very sensitive to input parameter values, especially precipitation input data. With several different sources of precipitation data now available, it is quite difficult to determine which source is most appropriate under various circumstances. We used several sources of rainfall data in this study including single gauge rainfall data located outside the watershed boundary, and next generation radar (NEXRAD) rainfall data with different corrections, to examine the impact of such sources on Soil and Water Assessment Tool (SWAT) model streamflow predictions tor a 50 km 2 watershed located in the coastal plain of Maryland. For a watershed of that size with annual average precipitation of 43 inches, at least 3 rain gauges within the watershed would reduce the percentage error in measured average watershed rainfall amounts to less than 23% (for 0.5 inch storm events). The larger the amount of storm rainfall the less error was associated with its measurement. Model simulation results indicated that distance and location of the single rain gauge located outside the watershed boundary has a significant impact in simulating hydrologic and water quality response of the watershed in the temperate region of Maryland. In the absence of a spatially representative network of rain gauges within the watershed, NEXRAD data produced more accurate estimates of streamflow than using single gage data. This study concludes that one has to be mindful of the source and methods of interpolation of rainfall data for input into hydrologic and water quality models if simulation accuracies are desired.