39Hydro Drought Prediction over the U.S. Using a Statistical Model of Groundwater and Land-Atmosphere Stable States

Prolonged drought has devastating societal impacts ranging from recreation to water and food supply and can ultimately result in significant financial losses. In a warming world, the frequency of seasonal- and multi-year drought episodes is likely to increase. Accurate drought prediction has the potential to reduce negative impacts and support societal resiliency by informing sustainable water-management and planning practices. Such predictions require a thorough understanding of the physical mechanisms that lead to the onset, persistence, and recovery of drought and the ability to replicate these processes in forecast models at appropriate time scales (weeks to months). Here we analyze the skill of a stochastic model of groundwater and atmospheric stable states. Stable states refer to consistent levels or conditions of groundwater over time, which can influence and be influenced by surface soil moisture, streamflow and atmospheric conditions (i.e. land-atmosphere interactions). This concept is used to create a forecast model based on the Coupling Stochastic Model (CSM; Roundy et al. 2014), which is a Markov-Chain model based on the persistence in L-A coupling regime but was updated to include a soil moisture and groundwater component to better capture the evolution of drought. A unique aspect of this model is that it can be applied globally using satellite remote sensing observations from SMAP (Soil Moisture Active Passive), AIRS (Atmospheric Infrared Sounder) and the GRACE (Gravity Recovery and Climate Experiment. This presentation focuses on assessing the skill of these drought predictions at sub-seasonal to seasonal time scales over the U.S. and discusses the limitations of a locally persistent statistical forecast model as well as ways to improve the forecasts. Despite the limitations, we argue that statistical models based on remotely sensed satellite observations provide an important tool for improved predictability and understanding of extreme events that complements physically based models.