@misc{tootchi2018mgwm, author={Ardalan {Tootchi} and Anne {Jost} and Agn\`{e}s {Ducharne}}, title={{Multi-source global wetland maps combining surface water imagery and groundwater constraints}}, year={2018}, doi={10.1594/PANGAEA.892657}, url={https://doi.org/10.1594/PANGAEA.892657}, organization={Sorbonne Universit\'{e}, Paris, France}, abstract={Many maps of open water and wetland have been developed based on three main methods: (i) compiling national/regional wetland surveys; (ii) identifying inundated areas by satellite imagery; (iii) delineating wetlands as shallow water table areas based on groundwater modelling. The resulting global wetland extents, however, vary from 3 to 21{\%} of the land surface area, because of inconsistencies in wetland definitions and limitations in observation or modelling systems. To reconcile these differences, we propose composite wetland (CW) maps combining two classes of wetlands: (1) regularly flooded wetlands (RFW) which are obtained by overlapping selected open-water and inundation datasets; (2) groundwater-driven wetlands (GDW) derived from groundwater modelling (either direct or simplified using several variants of the topographic index). Wetlands are thus statically defined as areas with persistent near saturated soil because of regular flooding or shallow groundwater. To explore the uncertainty of the proposed data fusion, seven CW maps were generated at the 15 arc-sec resolution (ca 500 m at the Equator) using geographic information system (GIS) tools, by combining one RFW and different GDW maps. They correspond to contemporary potential wetlands, i.e. the expected wetlands assuming no human influence under the present climate. To validate the approach, these CW maps were compared to existing wetland datasets at the global and regional scales: the spatial patterns are decently captured, but the wetland extents are difficult to assess against the dispersion of the validation datasets. Compared to the only regional dataset encompassing both GDWs and RFWs, over France, the CW maps perform well and better than all other considered global wetland datasets. Two CW maps, showing the best overall match with the available evaluation datasets, are eventually selected. They give a global wetland extent of 27.5 and 29 million km${^2}$, i.e. 21.1 and 21.6{\%} of global land area, which is among the highest values in the literature, in line with recent estimates also recognizing the contribution of GDWs. This wetland class covers 15{\%} of global land area, against 9.7{\%} for RFWs (with an overlap ca 3.4 {\%}), including wetlands under canopy/cloud cover leading to high wetland densities in the tropics, and small scattered wetlands, which cover less than 5{\%} of land but are very important for hydrological and ecological functioning in temperate to arid areas. By distinguishing the RFWs and GDWs globally based on uniform principles, the proposed dataset is believed to be useful for large-scale land surface modelling (hydrological, ecological and biogeochemical modelling) and environmental planning.}, type={data set}, publisher={PANGAEA} }