Sensitivity of convective precipitation forecasts to soil moisture and vegetation

Land surface properties play a major role in convective precipitation events through impacting the amount of surface evaporation which results in changes to near surface temperature and humidity. This study examines the effects of using soil moisture data from the North American Regional Reanalysis (NARR) and the Soil Moisture Ocean Salinity Satellite (SMOS) on short term weather forecasts using the Weather Research and Forecasting Model (WRF).;SMOS soil moisture data were compared to in-situ observations and it was found that although they captured the spatial variation in soil moisture, the actual measurements had a dry bias of roughly 0.10 m3/m 3. Large differences existed between the in-situ observations, even for probes only a few meters apart. Observations from different sensors within a SMOS footprint differed from each other by a larger amount than they differed from the SMOS retrieval. Removing the mean and normalizing the data brought the in-situ observations into better agreement with each other and with SMOS but they still contained substantial differences.;WRF sensitivity experiments demonstrated that changes to initial values of soil moisture resulted in no significant changes in precipitation. However, more of an impact was seen when the vegetation was changed, with barren vegetation yielding a substantial decrease in precipitation. Adding soil moisture resulted in significant changes to 2 m temperature and dewpoint relative to the control runs for each vegetation type. However, it was found that convective available potential energy and moist static energy change little, as the temperature and humidity impacts on these variables cancel each other out, which explain the limited precipitation response.;SMOS data resulted in no significant changes in precipitation forecasts but had some impacts on temperature and humidity forecasts. However, because these results were not seen in all cases, no definitive conclusions about the usefulness of SMOS for high resolution numerical modeling can be made at this time. These results provide major implications for future satellite missions such as Soil Moisture Active Passive showing that experiments using true data assimilation methods which give only partial weight to satellite data may also not provide significant improvements to weather forecasts.