Spatio-temporal analysis of drought using ground and remote sensing data

Abstract

The aim of this paper is the spatial and temporal analysis of droughts in a typical water limited Mediterranean basin, Pinios River Basin in Thessaly, Greece using satellite data and compare this analysis with the spatio-temporal analysis of droughts using ground station data. To achieve this, time series analysis with geoinformatics techniques and the calculation of drought indices is applied. The drought assessment methodology was developed for the determination of spatio-temporal drought analysis using ground and remote sensing data. The analysis was carried out for the common period from October 1981 to September 2002. The methodology was developed by estimating the meteorological drought indices, Standardized Precipitation Index (SPI) and Standardized Precipitation Evapotranspiration Index (SPEI) for various time scales, but also comparing them with the Surface Runoff Index (hydrological drought index).  Initially, the station monthly precipitation and temperature data were spatially distributed by using an MLR interpolation method.  Thus, a canvas of 487 pixels was created for the entire basin area.  Correlation analysis was performed to validate the results of the spatial distribution of monthly precipitation and temperature.  Precipitation satellite data by Climate Hazards Infrared Precipitation with Stations (CHIRPS) have been used. These global precipitation data have a spatial resolution of 0.05° (5 x 5 km), from 1981.  The monthly temperature was estimated by the ERA5 reanalysis data at the same spatial resolution. The Thornthwaite method was used for the estimation of potential evapotranspiration for both ground and satellite data. The values of the drought indices (SPI and SPEI) estimated using ground and remote sensing data have been compared for the whole period of analysis (1981-2002) and the results were deemed satisfactory. The Pinios River Basin experienced severe, extreme, and persistent droughts during the period from the late 1980s to the early 1990s and in early 2000s and the results of the analysis have been compared for these specific drought events. Furthermore, the drought indices of various time scales were compared for an unregulated sub-basin of Pinios River with the SRI index and the results indicated that a meteorological drought index (SPI and/or SPEI) at 6-month time scale correlates well with the SRI. Overall, the results of the analysis showed that the satellite data could be used for the estimation of drought indices with reasonable accuracy for the analysis of droughts in areas were ground meteorological data are not available. Keywords: Droughts; Standardized Precipitation Index, SPI; Standardized Precipitation Evapotranspiration Index, SPEI; Surface Runoff Index, SRI; CHIRPS; ERA5; Pinios River Basin