Comparison of standardized precipitation index (SPI) and vegetation health index (VHI) in drought monitoring in Isingiro District.

Abstract

Drought is a recurring and most complex weather-related natural phenomenon, affecting vast areas and communities around the world every year due to its slow onset and cascading effects. Droughts are common in Uganda’s cattle corridor, but there is limited information on their occurrence and severity. The magnitude and pattern of drought can be measured with various drought indices using remote sensing and meteorological data. Many indices are used to monitor drought events. However, different indices have different data requirements and applications. Hence, evaluating their applicability will help to characterize drought events and refine the development of effective drought indices. The characteristics of drought events from 2000 to 2020 were compared using Standardized Precipitation Index and Vegetation Health Index at a five-year interval. Drought was assessed using Landsat 8 OLI and 7 ETM temporal images based on Vegetation Health Index for the years; 2000, 2005, 2010, 2015 and 2020. In addition, SPI was utilized together with DrinC software and meteorological (monthly precipitation) data from the Uganda National Meteorological Authority to compute, determine and graphically illustrate trend in drought severity from 1990 to 2020 at various time scales (3,6 and 12 month). Vegetation Health Index and Standardized Precipitation Index were used to identify the temporal and spatial drought patterns as derived from the study of a 20-year time period. The analysis revealed that 2000 and 2015 were the driest years. Using SPI, the southeastern and northwestern parts of the Isingiro district are the areas that are susceptible to drought and while using VHI, it was the northwestern, southwestern and southwestern parts of Isingiro are the areas that are susceptible to drought. SPI works well when there is an even distribution of weather stations in the area. SPI-3 and VHI-2010 had the lowest correlation of 0.105 at the 3-month time scale and the highest correlation was in 2000(0.541) and 2015(0.659). VHI showed more areas affected by drought compared to SPI. The comparative analysis of the two indices indicated that VHI and SPI are highly correlated in the years 2000 and 2015. Drought analysis based on these indices showed that for drought assessment VHI can be used to depict drought condition of the study area more realistically. The results of this comparison show that 12-month scale of SPI and VHI have a higher correlation.