Statistics of multi‐year droughts from the method for object‐based diagnostic evaluation

Abstract

This study uses the method for object‐based diagnostic evaluation (MODE) technique to examine and compare the statistics of drought attributes over the upper Colorado River basin (UCRB). The drought objects are based on the standardized precipitation index (SPI) and the standardized precipitation evapotranspiration index (SPEI) on a 36‐month timescale (SPI36 and SPEI36, respectively). The drought indicators are calculated using monthly precipitation as well as minimum and maximum temperatures from the Precipitation–Elevation Regression on Independent Slopes Model data sets from 1948 to 2012. MODE uses paired object attributes such as centroid distance, orientation angle, area ratio, and intersection area and a combination of parameter thresholds to determine the number of objects identified and retained in the merging and matching process in the two fields. Using MODE run with convolution radius of zero (no smoothing) and an area threshold of four grid points, this study computes and analyses object statistics including centroid locations, areas, and intensity percentiles. Results of the analysis show that SPI36 produces more drought objects than SPEI36. Although the spatial patterns are roughly similar leading up to almost similar statistics of object attributes, such as locations of the object centroids, the SPI36 produces higher percentile intensity of drought objects than does SPEI36, which is clearly obvious in the 90th percentile intensity of drought objects. The largest difference between SPEI36 and SPI36 occurs in the area of drought objects during the early 2000s when the region experienced multi‐year drought resulting from increased warming of the atmosphere. This study demonstrates the use of MODE as a tool to evaluate and monitor drought event over the UCRB.