Significance of ELDAS soil moisture products for NWP

Abstract

The analysis of soil moisture for the initialization of numerical weather prediction (NWP) models is discussed subjected to the constraints imposed by an operational environment. Three different techniques are compared within the HIRLAM forecasting system. The first method is the HIRLAM default option based on optimal interpolation (OI) analysis with optimal coefficients analytically formulated. It makes use of 2-metre temperature and relative humidity errors. A second method is based on a simplified variational approach to assimilate also 2-metre observations of temperature and relative humidity. The estimate of tangent linear of the observation operator is obtained here from an extra integration of the numerical model. Finally, a third method is based on the assimilation of soil moisture by a parent (global) model and a posteriori corrected by the forecasted precipitation error. The soil moisture produced is then imported to the HIRLAM system.