Analyzing the Cycle of Flash Recovery-Flash Drought-Flash Recovery in 2015 over the South-Central U.S. (Invited Presentation)

Flash drought, defined as a rapid onset and intensification of drought, is characterized by abnormally warm and dry weather conditions that leads to the rapid depletion of soil moisture and rapid deteriorations in vegetation health from increased evaporative stress. Flash recovery, on the other hand, is characterized by a period(s) of intense precipitation, usually over the course of a few weeks, where drought conditions are quickly eradicated and may be replaced by waterlogged soils and flooding. Large sections of the south-central U.S. not only had both in the same season, but a very unusual sequence of flash recovery-flash drought-flash recovery during a six-month period between May and November in 2015. A detailed analysis of the soil moisture anomalies during the event was performed for three select sub-regions stretching from central Texas to Mississippi using five different combinations of different land surface models (LSM’s) and meteorological forcing datasets in the NASA Land Information Systems (LIS) framework. Preliminary results show decent spatial and temporal agreement in soil moisture anomalies between the five runs during the first flash recovery in May. However, there was significant disagreement between runs regarding the evolution of the flash drought that began in late July and came to an abrupt halt with a second flash recovery in late October. The disagreement was more significant between runs with the same land surface models but different forcing datasets (i.e. the North American Land Data Assimilation Systems (NLDAS) vs. Global Data Assimilation System (GDAS)) than between runs with different land surface models (Noah 3.6 vs. Noah-MP/Noah-MP with dynamic vegetation) but same forcing datasets. Significant differences were also noted in the depiction of flash drought local intensification between runs with different land surface models but same meteorological forcing. No run exactly matched the evolution as depicted by the U.S. Drought Monitor but it should also be noted that each run had some value that shouldn’t be discounted. In this presentation we will discuss the evolution of this unique flash recovery-flash drought-flash recovery cycle, the possible causes for the differences depicted by the five runs, and some recommendations on optimal model configurations for future process studies such as this.