Eta model estimated land surface processes
and the hydrologic cycle of the Mississippi basin

E. Hugo Berbery

University of Maryland

Abstract:
This presentation will discuss the water and surface energy budgets over the Mississippi basin and sub-basins resulting from NCEP's Eta model forecasts. It will also discuss the relation between surface states (soil moisture) and other variables that affect the surface energy balance, potentially interacting with precipitation processes.

On a 7-yr average, the Eta model 12-36 h forecast precipitation averaged over the Mississippi basin differs from the observed precipitation by 2%, while the estimate of evaporation computed as a residual of the water balance equation differs by 5% from the evaporation estimate resulting of a dataset of land surface fluxes prepared with the macroscale hydrologic Variable Infiltration Model (VIC). Notably, the long term average of moisture flux convergence is 0.49 mm day-1 over the Mississippi basin, while streamflow observations at Vicksburg average 0.50 mm day-1. This agreement is a strict test of the quality of the hydrologic cycle estimates, therefore these are promising results for estimates of the water cycle from the Regional Reanalysis currently being produced.

Sub-basins of the Mississippi have diverse land surface-atmosphere interactions as inferred from monthly averages. In the western half of the Mississippi basin, feedbacks can be described as follows: increased soil moisture is associated with a slight increase of net radiation at the surface; latent heat also increases with soil moisture while sensible heat decreases, resulting in an almost linear increase of the evaporative fraction. The overall results support the concept of a positive feedback in which increased soil moisture affects surface fluxes in such a manner that increased precipitation results. However, toward the east (e.g., the Ohio basin), there are no well defined land surface-atmosphere interactions, suggesting that other effects, like the advection of moisture, may be more relevant for precipitation processes.