Rapid environmental change in the Arctic raises concerns about the future of wetland carbon dynamics. Current classifications may not capture the diverse environmental conditions affecting these fluxes. This study aims to fill this gap by identifying distinct wetland clusters using a synthesis of field measurements (CO2 and CH4 fluxes, soil moisture) and remote sensing and other geospatial data (snow season length, annual precipitation, growing degree days, normalized difference vegetation index, and potential peat depth). Our dataset includes 862 individual measurements of tundra and forest-tundra wetlands from 86 data sources covering the period 1988 - 2023 for a variety of Arctic regions, with dominant wetland types including fens and polygonal bogs. Clusters for CO2 and CH4 fluxes mostly did not overlap, indicating different controls on these two carbon fluxes. CH4 flux clusters depended on local moisture conditions, potential soil depth, and vegetation characteristics, while CO2 flux clusters varied significantly with climatic conditions and vegetation characteristics. Notably, no single wetland type from the current classification formed a distinct cluster in either carbon flux. This study highlights the importance of high spatial resolution soil and vegetation data for accurate scaling of wetland CH4 fluxes. However, this may be less critical for the upscaling of wetland CO2 fluxes.