The Influence of Satellite-based Estimates of Precipitation on Modeled Land Surface States and Fluxes over the North American Monsoon Region

The Enhanced Observing Period of the North American Monsoon Experiment was conducted from June through September of 2004. During this period, and in subsequent years, several tower flux sites collected data on land atmosphere exchanges of energy and moisture from key representative land cover types. In this presentation we evaluate the performance of two leading land surface parameterizations (Noah and CLM) in their ability to simulate observed land surface-atmosphere exchanges from these flux tower sites. In addition to available surface data, we also analyze the impact of using several different operational remotely- sensed estimates of precipitation as forcing data. Previous work has shown a wide range in precipitation estimates from space-borne platforms over the N. American monsoon region and here we examine how this uncertainty in precipitation estimates propagates through simulated state and flux variables. Our findings suggest that un-critical use of such estimates can yield significant biases in simulated estimates of runoff, soil moisture and evapotranspiration. We also compare our findings in the context of operational re-analysis products that are being generated over N. America. A brief explanation for the error structures in the different precipitation products is provided and we conclude the presentation with a set of recommendations for using remotely-sensed precipitation products for land surface modeling studies in the monsoon region.