Sensitivity of surface albedo derived from METEOSAT data to drought episodes in a semi-arid region (M’Sila, north center of Algeria)

Abstract

Drought is a natural phenomenon that affects ecological processes, land productivity, and human life. The phenomenon is slow to settle down and its impacts are not instantaneous, requiring an early monitoring of their signs of appearance. In order to track the effect drought on the land surface dynamics by means of surface albedo changes, we intend to exploit the VIS-channel reference image from the METEOSAT satellite data, which is an associated product of the new version of the Infrared Difference Dust Index (IDDI). The surface albedo is retrieved from the calibrated VIS-channel reference radiance, atmospherically corrected by the radiative transfer code 6S (Second Simulation of a Satellite Signal in the Solar Spectrum). In this study, we inspect the sensitivity of monthly surface albedo to drought events over a semi-arid region (M’Sila, center of northern Algeria) during the period covered by three METEOSAT satellites (5, 6, and 7). Two drought indices, SPI (Standardized Precipitation Index) and SPEI (Standardized Precipitation Evapotranspiration Index), were calculated at multiple timescales and were correlated with surface albedo over the study region. The results show that the correlation of surface albedo with the two drought indices at the timescale of 1 month (current month) is not necessarily the best indicator of the installation of a drought event. Additionally, we find that the sensitivity of surface albedo to drought indices at a medium timescales class (average of correlation coefficient at 6, 9, and 12 months), is more sensitive to SPEI index (principally 6-month timescale), during the growing season (April to September). It is also more sensitive to SPI index (principally 12-month timescale), during the supplying and charging period of water resources (October to March).