Coarse land cover datasets bias Arctic-Boreal wetland methane budgets

Accurate methane (CH₄) emission estimates from Arctic and boreal wetlands are essential for reducing global budget uncertainties but are hindered by poorly constrained wetland distribution and classification. We assessed how land cover map resolution and thematic detail influence these estimates. Using very high spatial resolution land cover maps (≤2.5 m) with five to seven harmonized classes and 4-50% wetland coverage, we estimated CH₄ emissions across seven Arctic and boreal sites in North America and Eurasia. Resampling to coarser resolutions (up to 5 km) revealed that CH₄ flux estimates remained within 13% error when resolution was ≤25 m pixel size. At resolutions coarser than 1 km, four of seven sites shifted from net CH₄ source to sink, due to misrepresentation of wetland extent in heterogeneous landscapes with small, fragmented wetlands. Thematic detail also proved critical, as fens-high CH₄ emitters-were disproportionately underrepresented in coarse (>1 km) maps relative to other wetland types. We also show that existing global or circumpolar land cover maps tend to misrepresent wetlands, either overlooking smaller features or overestimating coverage in wetland dominated areas. Our findings indicate that coarse-scale land cover datasets are unsuitable for estimating CH₄ budgets in these regions, where high spatial resolution and biogeochemically relevant land cover classes are essential for reliable CH₄ emission upscaling.