Abstract Drained and cultivated grasslands on peat soils behave as a significant source of greenhouse gasses by oxidation. However, the lack of empirical estimates of carbon losses from peatlands with adequate spatial and temporal resolution has forced researchers to rely on process‐based model approximations to make quantitative, regional‐ or national‐scale estimations. Here we use satellite‐based synthetic aperture radar interferometry to estimate the land motion per parcel with a daily resolution, discriminate a reversible and an irreversible component, and convert this to an upper bound of CO2 ${text{CO} }{2}$‐equivalent emissions over the western part of the Netherlands. We find an upper bound of 21.5 tCO2 ${text{tCO} }{2}$‐eq/ha/yr, corresponding to a total regional output of 2.3 MtCO2 ${text{MtCO} }_{2}$‐eq/yr, or approximately 1.3% of the entire greenhouse gas emissions of the Netherlands in 2019. The method also allows us to provide estimates for future emissions as well as evaluate the efficacy of installed subsidence mitigation measures.