Multi-source global wetland maps combining surface water imagery and groundwater constraints

Abstract

Abstract. Many maps of open water and wetlands have been developed based on three main methods: (i) compiling national and regional wetland surveys, (ii) identifying inundated areas via satellite imagery and (iii) delineating wetlands as shallow water table areas based on groundwater modeling. However, the resulting global wetland extents vary from 3 % to 21 % of the land surface area because of inconsistencies in wetland definitions and limitations in observation or modeling systems. To reconcile these differences, we propose composite wetland (CW) maps, combining two classes of wetlands: (1) regularly flooded wetlands (RFWs) obtained by overlapping selected open-water and inundation datasets; and (2) groundwater-driven wetlands (GDWs) derived from groundwater modeling (either direct or simplified using several variants of the topographic index). Wetlands are statically defined as areas with persistent near-saturated soil surfaces because of regular flooding or shallow groundwater, disregarding most human alterations (potential wetlands). Seven CW maps were generated at 15 arcsec resolution (ca. 500 m at the Equator) using geographic information system (GIS) tools and by combining one RFW and different GDW maps. To validate this approach, these CW maps were compared with existing wetland datasets at the global and regional scales. The spatial patterns were decently captured, but the wetland extents were difficult to assess compared to the dispersion of the validation datasets. Compared with the only regional dataset encompassing both GDWs and RFWs, over France, the CW maps performed well and better than all other considered global wetland datasets. Two CW maps, showing the best overall match with the available evaluation datasets, were eventually selected. These maps provided global wetland extents of 27.5 and 29 million km2, i.e., 21.1 % and 21.6 % of the global land area, which are among the highest values in the literature and are in line with recent estimates also recognizing the contribution of GDWs. This wetland class covers 15 % of the global land area compared with 9.7 % for RFW (with an overlap of ca. 3.4 %), including wetlands under canopy and/or cloud cover, leading to high wetland densities in the tropics and small scattered wetlands that cover less than 5 % of land but are highly important for hydrological and ecological functioning in temperate to arid areas. By distinguishing the RFWs and GDWs based globally on uniform principles, the proposed dataset might be useful for large-scale land surface modeling (hydrological, ecological and biogeochemical modeling) and environmental planning. The dataset consisting of the two selected CW maps and the contributing GDW and RFW maps is available from PANGAEA at https://doi.org/10.1594/PANGAEA.892657 (Tootchi et al., 2018).

The role of northern peatlands in the global carbon cycle for the 21st century

“…Based on the rationale ofStocker et al (2014), a cost-efficient version of TOPMODEL was adapted in our model to reduce the computational cost; an asymmetric sigmoid function is used to approximate the empirical relationship between the flooded fraction of a grid cell and its mean water table position calculated at each time step by the land surface model. The asymmetric sigmoid function is calibrated for each grid cell (Supporting InformationFigure S1b,c shows the simulated present-day peatland area fraction and peat soil carbon density with the function being globally uniform, could be contrasted with simulations in the main manuscript which used grid-by-grid calibrated functions) so that the simulated maximum flooded area fraction of the grid cell matches with the present-day wetland areas that are regularly inundated or subject to shallow water tables after excluding lakes and permanent water bodies(Tootchi, Jost, & Ducharne, 2019).Then a scheme of peatland initiation and development is included in the land surface model, to distinguish areas that are suitable from the simulated wetland extent to become a peatland. Conditions required for peat inception and persistence are; (a) a minimum flooding duration over a long period (at least Num flooded months during 30 years, with Num being the number of growing season months during the 30 years); (b) a positive summer water balance; and (c) a long-term positive carbon balance.…”

ORCHIDEE MICT-LEAK (r5459), a global model for the production, transport, and transformation of dissolved organic carbon from Arctic permafrost regions – Part 1: Rationale, model description, and simulation protocol

“…Such a floodplain area is represented as a fraction of a grid cell with the maximum extent of inundation, termed the "potential flooded area" being predefined from a forcing file (Tootchi et al, 2019). Here, the DOC pools that are already being produced in these inundated areas from litter and SOC decomposition in the first five layers of the soil column are directly absorbed by the overlying floodwaters.…”

“…The spatial areas that are available for potential flooding are predefined by an input map originally based on the map of Prigent et al (2007). However, for this study, we used an alternative map of the "regularly flooded areas" derived from the method described in Tootchi et al (2019), which in this study uses an improved input potential flooding area forcing file specific to the Lena basin that combines three high-resolution surface water and inundation datasets derived from satellite imagery: GIEMS-D15 (Fluet-Chouinard et al, 2015), which results from the downscaling of the map of Prigent et al (2007) at 15 arcsec (ca 500 m at the Equator); ESA-CCI land cover (at 300 m ∼ 10 arcsec); and JRC surface water at 1 arcsec (Pekel et al, 2016). The "fusion" approach followed by this forcing dataset stems from the assumption that the potential flooding areas identified by the different datasets are all valid despite their uncertainties, although none of them are exhaustive.…”

“…Rapid melting of snow and soil or river ice during spring freshet (May-June) drives intense seasonal discharge, with peaks often 2 orders of magnitude (e.g. Van Vliet et al, 2012) above baseflow rates (Fig. 1d).…”

“…High suspended particle loads in river water as they approach the mouth (Heim et al, 2014) cause lower light availability and water albedo and hence higher temperatures (Bauch et al, 2013;Janout et al, 2016), which can affect the nearshore sea ice extent, particularly in spring (Steele and Ermold, 2015). Volumes of riverine freshwater and total energy flux (Lammers et al, 2007) are expected to increase with warmer temperatures, along with an earlier discharge peak (Van Vliet et al, 2012. In doing so, freshwaters may in the future trigger an earlier onset of ice retreat (Stroeve et al, 2014;Whitefield et al, 2015) via a feedback between freshwater albedo, ice melt, and seawater albedo amplified by in-termediary state variables such as water vapour and cloudiness (Serreze and Barry, 2011).…”