Reconciling differences in observed annual methane growth rates

Satellite and surface observations of the annual growth rate of atmospheric methane (MGR) often show notable differences, which are typically interpreted as inter-platform discrepancies. The surface network is conventionally considered the more accurate benchmark. Using a chemical transport model (GEOS-Chem) with precisely known methane budget terms, we demonstrate that these differing MGR estimates can be inherently consistent and do not necessarily reflect observational error. By sampling the model to create pseudo-observations, we show that satellite-derived MGR aligns more closely with the model's "true" global growth rate than the surface-network-derived MGR at monthly-to-annual scales. This superior performance stems from the satellite's more comprehensive spatial coverage. Perturbation experiments reveal that the sparse and uneven distribution of surface sites systematically biases global burden estimates in response to short-term, regional emission changes—underestimating signals from poorly sampled regions (e.g., tropical Africa) and overestimating those from well-sampled regions (e.g., East Asia). In contrast, satellite observations provide a more robust estimate of such perturbations, although gaps in coverage can still introduce bias. We conclude that for monitoring the global methane growth rate, satellite observations offer greater representativeness and accuracy than the surface network, reconciling apparent discrepancies and redefining the hierarchy of observational constraints for global-scale methane variability.