A remote sensing-based modeling study of land-atmosphere fluxes during climatic extremes in the Mississippi River Basin

Climate and forecast models require mathematical representations of land-atmosphere interactions. Such representations depend heavily on the specification of key biophysical parameters. These parameters are highly heterogeneous in space and time, making it difficult to prescribe them accurately. Data anticipated from the EOS generation of sensors are expected to contribute significantly to overcoming these difficulties because of their global coverage and high spatial and temporal information content. As a precursor to the use of actual EOS data, the objective of the research reported here was to establish a methodology for using EOS-era data to evaluate and improve the representation of the land surface in a regional climate model.; Regional model simulations of the 1988 drought and 1993 flood events in the Upper Mississippi River Basin were used to supply atmospheric and standard land surface conditions to the Biosphere-Atmosphere Transfer Scheme (BATS). Land surface parameter data sets (land-cover type, leaf area index (LAI), albedo, roughness length) were developed from existing remotely-sensed sources and by applying EOS-like algorithms to remotely-sensed data at relatively high spatial resolution. These data were used to evaluate standard model parameters and in model simulations to assess their impact on land-atmosphere fluxes simulated by BATS.; In general, land surface conditions specified by the remotely-sensed data were temporally and spatially heterogeneous than those prescribed by BATS. These data had significant impacts on simulated surface fluxes. When parameters were incorporated individually, BATS was found to be particularly sensitive to the remotely-sensed LAI. The simultaneous incorporation of all parameters resulted in an increase in the Bowen ratio of 128%. Differences in moisture availability between the two years played a role in determining sensitivity.; It was also shown that empirical corrections and assumptions implicit in the satellite products had a large impact on the remotely-sensed parameters and thus on the simulated surface fluxes. Improved algorithms and atmospheric data in the EOS era are expected to reduce the current uncertainty associated with these parameters. The results of this investigation show that EOS-era data sets will be both useful and necessary for evaluating and improving the specification of the land surface in climate models.