Supporting Water Security and Climate Change Resilience by Quantifying the Uncertainties in Global Surface Water Extent Estimates.

Changing climate and anthropogenic disturbances are affecting global surface water availability. Recent abundance in satellite Earth observations have assisted in remotely monitoring surface water bodies. However, there is no baseline to quantify the inconsistencies across the existing Global Surface Water Extent (GSWE) datasets and the uncertainties therein. This study is the first to fill this knowledge gap by comparing two latest GSWE datasets ‒ the 30-m Joint Research Centre (JRC) Global Surface Water Occurrence and 10-m European Space Agency (ESA) WorldCover, derived from the Landsat and Sentinel satellites respectively, as well as a GSWE dataset called HydroLAKES derived from the HydroSHEDS global merged hydrography. Results suggest that, on the surface, these GSWE data may appear to be similar, however, substantial inconsistencies exist in total count, mean area, and total area across these datasets. For example, there exists a 20 times variation between the number of water bodies captured by ESA and HydroLAKES datasets across the U.S., despite the mean area of water bodies in HydroLAKES being 50% more than that for ESA water bodies. This study will provide critical new information to understand the uncertainties in what we know about global surface water availability and how we plan to sustain the available water against global environmental changes.