Landsat-based remote sensing of surface water dynamics in southern African wetlands in drylands from 1986 to 2022

Abstract

Wetlands in drylands represent distinct ecosystems characterized by unique assemblages of fauna and flora that coexist in a fluctuating hydrological regime. These habitats provide essential resources, including water and vegetation, which support fauna and thus concentrate a complex web of ecological interactions that rely on the presence of water, especially during the dry season. Understanding the alternating dynamics of recharge and drying is therefore critical, as these processes directly affect the ecological interactions that depend on water availability. To better understand these dynamics, we developed a Landsat-based remote sensing methodology using a 37-year-long time series of images to assess the frequency of water presence in known wetlands in drylands in Hwange National Park, Zimbabwe, in relation to climate data obtained from ERA5. Annual and monthly frequency of water presence was calculated using a thresholding single water index based on the modified normalized detection water index. We assessed inter- and intra-annual variations in water presence. This methodology provides a month-by-month assessment of water presence and anomalies from the average between 1986 and 2022. The filling and drying of wetlands showed a temporal lag of 2 months between precipitation and water presence, thus allowing the assessment of the ecological drought intensity in the dry season. Overall, the results allow us to evaluate the drying and filling dynamics of wetlands in drylands as well as their surface variability, giving crucial insights into the water availability dynamics and their resilience to climate forcing and variations.