Methane (CH4) is a potent greenhouse gas that has been increasing rapidly in the atmosphere since 2007, following a stable period during 2000-2006. It is essential to understand its budget, including changes from both emissions and sinks. The potential sources and sinks are relatively well understood; however, their relative contributions to atmospheric CH4 still remain poorly known. For example, the scientific community is still debating what is responsible for the post-2007 increase in atmospheric CH4. The top-down approach (inverse modeling) is often employed by the scientific community to infer surface fluxes using atmospheric CH4 measurements, and inverse models that assimilate both CH4 concentrations and atmospheric δ13C measurements point to a dominant contribution to the rapid increase from microbial emissions. Recent research using box modeling has shown that atmospheric δD (proportional to the D/H ratio of CH4) has the potential to further improve our understanding of methane sources and sinks. In this presentation, we will further explore the utility of D using three-dimensional simulations and comparisons with observations between 2011 and 2022. Atmospheric δ13C has proven to be a useful tracer in distinguishing between fossil fuel and microbial emissions. On the other hand, the lack of spatio-temporal signatures of CH4 sources gives rise to considerable uncertainty in our understanding of the global CH4 budget. Moreover, wetland emissions exhibit large spatio-temporal variability with a significant uncertainty in inverse estimates. Uncertainty in the budget also arises from a lack of understanding of the CH4 sink due to OH, Cl, and O1D. Atmospheric δD fractionates due to reaction with OH differently than δ13C, so atmospheric δD could be helpful to constrain the atmospheric CH4 sink. Additionally, the D signature of most emissions varies strongly with latitude, providing a different constraint than 13C. We will compare globally distributed atmospheric δD measurements to simulations forced by surface fluxes calculated using the CarbonTracker-CH4 data assimilation (inverse modeling) system, as well as newly constructed maps of δD source signatures. By construction, the CarbonTracker fluxes (CT2025) already agree relatively well with global observations of CH4 and δ13C. However, CarbonTracker fluxes are tied to assumed methane loss rates and δ13C fractionation factors as well as prior spatial distributions of surface fluxes, such as tropical wetland extent. The series of δD simulations and comparisons with observations we will present allow us to test these assumptions against δD observations. Here, we examine several emission scenarios to simulate atmospheric δD. Preliminary results using CT2025 fluxes show that simulated atmospheric δD generally agrees with observations, indicating the overall quality of the CarbonTracker inversion results. However, the CT2025 fluxes overestimate observed δD at Northern Hemisphere sites and are comparable at Southern Hemisphere sites. In contrast, the “Ch4only” scenario, which does not include 13C data constraints, greatly underestimates observed atmospheric δD. “Mic+” and “TropWL+” scenarios, which have enhanced microbial emissions relative to CT2025, show improved agreement with observed δD in the Northern Hemisphere, including reproducing the observed long-term global means and latitudinal gradients of CH4, δ13C. These D results suggest a potential avenue for improvement in the standard CarbonTracker fluxes. All simulations show minimum atmospheric δD one month earlier than observations in the Southern Hemisphere; the model also exhibits a smaller seasonal cycle amplitude than observations in the Northern Hemisphere, indicating the possibility that the seasonality of CH4 sinks in our standard inversions can also be improved. Our eventual goal is to constrain the global CH4 budget by simultaneously fitting CH4 mole fraction, atmospheric δ13C, and δD measurements in an inverse modeling framework. In this study, we will demonstrate the extent to which atmospheric δD can be a potential tracer to refine our present understanding of the global CH4 budget.