Regional model simulations of the 2008 drought in southern South America using a consistent set of land surface properties

Abstract

This work discusses the land surface-atmosphere interactions during the severe drought that took place in 2008 in southern South America. The drought was among the most severe in the last fifty years both in terms of intensity and extent. Once precipitation returned to normal values, it took about two months for the soil moisture content and vegetation to recover. The land surface effects were examined by contrasting long term simulations using a consistent set of satellite-derived annually varying land surface biophysical properties against simulations using the conventional land cover types in the coupled system Weather Research and Forecasting Model/Noah Land Surface Model (WRF/Noah). The new land cover data set is based on ecosystem functional properties that capture changes in vegetation status due to climate anomalies and land use changes.The results show that the use of realistic information of vegetation states enhances the model performance reducing the precipitation biases over the drought region as well as over areas of excessive precipitation. The precipitation bias reductions are traced back to the corresponding changes in greenness fraction, leaf area index, stomatal resistance and surface roughness. The simulation of temperature shows a larger bias over the domain´s central part, which is attributable to a doubling of the stomatal resistance that reduces the evapotranspiration rate and leads to a temperature increase. However, the temperature pattern using the novel data set shows improvements towards the eastern part of the domain. The overall results suggest that an improved representation of the surface processes contributes to the predictability of the system.