Pattern Characterization of Meteorological Drought Using Multivariate Drought Index Over Mirzapur in Middle Gangetic Plains of India

Abstract

Droughts and floods have been occurring at a higher frequency in recent decades. The rapid transition between them magnifies the socio-economic consequences of these catastrophes relative to the effects of the individual occurrences of the extreme event. This study examines the temporal variability of meteorological drought and wet event characteristics occurring over Mirzapur (Uttar Pradesh), India. The Standardized Precipitation Evapotranspiration Index (SPEI) is applied to monthly water balance at scales 3, 6, 9, and 12 months to estimate the meteorological drought and wet events from 1971 to 2018. Drought and wet event characteristics such as the number of drought/wet events, severity, duration, and intensity are estimated using run theory over SPEI output. While characterizing the pattern of trends over the historical time period, variable-sized cluster analysis (VSCA) allows the detection of multiple change points as opposed to the Mann-Kendall (MK) test, which produces a monotonic trend for the entire time period. The VSCA technique accounts for drought variability and depicts the pattern’s evolution across the period under consideration. Station-scale drought data from Mirzapur, Uttar Pradesh (UP), India, were used in the procedure. VSCA allows for the detection of many change points while describing the pattern of drought trend throughout a historical period, as opposed to the usual Mann-Kendall (MK) test, which provides a monotonic trend for the whole time. As a result, VSCA demonstrated the MK test compatibility.

Keywords

• multivariate drought index, wet event, spei, run theory, pattern characterization, variable-sized cluster analysis