Using SMAP data to improve drought early warning over the US Great Plains

Abstract:

A drought prone region such as the Great Plains of the United States (US GP) requires credible and actionable drought early warning. Such information cannot simply be extracted from available climate forecasts because of their large uncertainties at regional scales, and unclear connections to the needs of the decision makers. In particular, current dynamic seasonal predictions and climate projections, such as those produced by the NOAA North American Multi-Model Ensemble experiment (NMME) are much more reliable for winter and spring than for the summer season for the US GP. To mitigate the weaknesses of dynamic prediction/projections, we have identified three key processes behind the spring-to-summer dry memory through observational studies, as the scientific basis for a statistical drought early warning system. This system uses percentile soil moisture anomalies in spring as a key input to provide a probabilistic summer drought early warning. The latter outperforms the dynamic prediction over the US Southern Plains and has been used by the Texas state water agency to support state drought preparedness. A main source of uncertainty for this drought early warning system is the soil moisture input obtained from the NOAA Climate Forecasting System (CFS). We are testing use of the beta version of NASA Soil Moisture Active Passive (SMAP) soil moisture data, along with the Soil Moisture and Ocean Salinity (SMOS), and the long-term Essential Climate Variable Soil Moisture (ECV-SM) soil moisture data, to reduce this uncertainty. Preliminary results based on ECV-SM suggests satellite based soil moisture data could improve early warning of rainfall anomalies over the western US GP with less dense vegetation. The skill degrades over the eastern US GP where denser vegetation is found. We evaluate our SMAP-based drought early warning for 2015 summer against observations.