Characterization of drought in Texas using NLDAS soil moisture data

From June to August 2011, Texas experienced the hottest summer ever recorded in the history of the United States, and the state suffered a water shortage that made its vulnerability to drought painfully plain. This disaster sparked new interest in methods of defining drought severity, especially with regard to the variation of soil moisture levels. This thesis assesses the suitability of information from the North American Land Data Assimilation System (NLDAS), an assemblage of land surface models forced with observations data, for quantifying soil moisture levels in Texas.

The potential for combining NLDAS data with the Soil Survey Geographic (SSURGO) Database’s available water capacity data is explored. It is discovered that because NLDAS is a hydrological model and SSURGO an agricultural dataset, they employ different definitions of soil moisture storage. Moreover, the temporal variation of soil moisture levels in the SSURGO polygons cannot be inferred from NLDAS data due to the vastly different spatial scales of the two datasets. A relative measure of soil saturation from 0–100% is developed instead and determined to be a more useful indicator of drought than the soil moisture level itself. Calculated solely from NLDAS data, it is used to map the severity of drought in Texas, with the results displayed at the county scale.

The temporal variation in soil moisture storage across the state is compared with variations in the gravity anomaly measured by NASA’s Gravity Recovery and Climate Experiment (GRACE) satellites and variations in Texas surface water reservoir levels, both of which are physical measurements of water storage changes. This analysis suggests that the NLDAS data, which is derived from a land surface model, accurately describes subsurface moisture variations. Also, the GRACE gravity anomaly data reveals that during the 2011 drought, the total water storage in Texas was approximately 100 cubic kilometers less than normal. NLDAS data indicates that more than 50% of this deficit was due to losses from the top one meter of the state’s soils.