Using Remotely Sensed Precipitation Information and Vegetation Observations for Early Drought Detection and Near-Real Time Monitoring on a Global Scale

This study evaluates the feasibility of using satellite precipitation data from the REDR program (CMORPH-CDR) to detect and monitor drought on a global scale. Monthly and daily Standardized Precipitation Indexes (SPI) were implemented and computed over various time scales. Results indicated that both monthly and daily SPIs presented the same timing and area for the major droughts episodes over the continental United States as well as for selected drought events around the globe. The SPI is evaluated primarily over CONUS where long-term drought monitoring products based on in-situ data exists such as the United States Drought Monitor (USDM), the nClimGrid derived SPI, and the WestWide Drought Tracker (WWDT). Showcases of selected severe drought events were used for validation (1998-2004 western US, 2006-2007 SE US, 2010-2012 TX-MX , 2012 summer Midwestern US). Following the assessment metrics in the NIDIS Drought Task Force (DTF) Protocol, each drought product is evaluated on the basis of its ability to estimate the onset and recovery, duration and severity, probability of drought condition, and the value given at the observed period. In addition to meteorological droughts, we also provide an assessment of drought metrics derived from the vegetation response as measured by the AVHRR Normalized Difference Vegetation Index CDR (NDVI-CDR). A global climatology was created based on data from 1981-2010, and anomalies are examined as drought indicators to compare to the SPI performance of estimating severe drought events. This effort looks forward to developing a hybridized approach for drought detection by incorporating both near-real time precipitation and near-real time vegetation remotely sensed information. Finally, we will present an interactive visualization tool that will allow easy comparison of the results for the selected drought events. Using geographic information system (GIS) technologies, monthly gridded SPI results were intersected with geopolitical boundaries (i.e., US States and Countries) to compute regional mean and counts of threshold exceedance. The counts were then used to evaluate spatial percentages based on the total number of grid cells within each area. These analyses provide a regionally-based timeseries of SPI results that can be updated in near-real time.