Evaluating Multi-Year Drought Events for the Contiguous U.S. in Contemporary and Scenario Climates using Object-Oriented Analysis
We identify drought events over the contiguous U.S. using the Method for Object-Based Diagnostic Evaluation (MODE) Time Domain (MTD). MTD uses object-based techniques to track 3-dimensional space-time objects, which we analyze here as drought events. This event-based analysis complements analyses that yield maps and other statistics of change in drought indices.
We use simulated daily precipitation from 34-members of the CESM Large Ensemble (CESM-LE) for four time periods: contemporary (1981 - 2010) and three projected periods under RCP 8.5 (2011 - 2040, 2041 - 2070, and 2071 - 2099). We use CRU and PRISM observational data to evaluate the quality of the CESM-LE simulations. From daily precipitation, we compute the 36-month Standardized Precipitation Index (SPI) and use it in MTD to identify multi-year drought events. SPI is a widely used index to characterize drought at different time scales. We use SPI thresholds to analyze drought events for six analysis regions covering the contiguous U.S. that were chosen to evaluate model performance in contemporary climate and to analyze regional projected changes in multi-year drought events.
For all regions, the CESM-LE results tend to fall between the two observational results. Drought changes toward the end of 21C, except in the central U.S., but the frequency and duration of multi-year droughts decreases across most of the U.S. The exception is the southwest U.S., where drought-event frequency changes little, however the area, duration, and intensity of multi-year droughts all increase.