Assessing the time variability of GIEMS-2 satellite-derived surface water extent over 30 years

Abstract

Inland waters, especially wetlands, play a crucial role in biodiversity, water resources and climate, and contribute significantly to global methane emissions. This study investigates the seasonal and inter-annual variability of the 0.25° × 0.25° surface water extent (SWE) from the Global Inundation Extent from Multi-Satellites (GIEMS-2) extended to a 30-year time series (1992–2020). Comparison with MODIS-derived SWE, CYGNSS-derived SWE and the Global Lakes and Wetlands Database (GLWD) shows consistent spatial patterns globally and over 10 different basins, although there are discrepancies in extent, partly due to different resolutions of the initial satellite observations. Strong cross-correlation (>0.8) in seasonal variability is observed when comparing GIEMS-2 with MODIS, CYGNSS and river discharge in most of the basins studied. Encouraging similarities were found in the inter-annual variability in most basins (cross-correlation >0.6) between GIEMS-2 and MODIS over 20 years, and between GIEMS-2 and river discharge over long time series, including over the Amazon and the Congo basins. These results highlight the reliability of GIEMS-2 in detecting changes in SWE in different environments, especially under dense vegetation, making it a valuable resource for calibrating hydrological models and studying global methane emissions.