Tracking Seasonal Precipitation's Dependence on Root Zone Soil Moisture Using Regional Reanalysis

Components of the hydrologic system such as streamflow, snowpack, soil moisture, and reservoir levels respond to changes in precipitation on timescales of weeks and months . As such, providing optimal early warning of drought onset requires outlooks for precipitation well beyond the range of a 7-day numerical weather model forecast. Marginal improvements can be made in seasonal drought outlooks by considering root zone soil moisture’s ability to alter evaporation levels, and in turn, precipitation. In this study North American Land Data Assimilation Systems (NLADS) was used to find regions within the central United States where root zone soil moisture is most strongly and positively correlated with subsequent seasonal precipitation over the 1985-2014 period of record. Correlation analysis reveals a migrating maximum region in the positive relationship between standardized root zone soil moisture anomalies and standardized precipitation indices for the 90 days to follow. In this region, root zone soil moisture explains 15-25% of the variance in seasonal precipitation. The maximum resides in New Mexico and western Texas when using soil moisture to predict precipitation for the March, April, and May season. The maximum migrates north-northeastward to northern Colorado, Wyoming, Nebraska, and Iowa by the June, July, and August season. The signal dissipates in the early fall. The observed correlation maximum appears to track locations where transitional zones between solar energy and soil moisture-limited evaporation regimes will manifest climatologically. These results from largely observation-driven reanalysis data are consistent with what has been seen in land-atmosphere coupling experiments. Furthermore, they provide additional evidence for the places in the United States where soil moisture is a relevant drought early warning tool at a given time of year.