Evaluating the Influence of Stratospheric Variability on Methane Total Columns

Atmospheric methane is a major driver of tropospheric oxidant budgets and a potent radiative forcing agent, and the trajectory of its atmospheric abundance will have significant consequences for the climate. However, studies that aim to determine which of its sources have produced the global trends observed over the last two decades disagree greatly. One prevailing approach to constraining the relative magnitudes of methane fluxes optimizes scales emissions estimates determined by bottom-up inventories based on the mismatch between methane concentrations resulting from a given chemical transport model (CTM) and mole fractions measured remotely and in situ at various scales. The ability of CTMs to explain seasonal to decadal trends in methane distributions thus has significant consequences for the validity of posterior emissions. The treatment of stratospheric chemistry and dynamics is especially important to consider in comparing modeled methane to total column dry-air mole fractions (DMFs). To assess the response of the measurement-model mismatch to modeled stratospheric variability, we compare total and tropospheric partial column DMFs from the Total Carbon Column Observing Network (TCCON) to vertical profiles derived from the GEOS-Chem CTM. Comparing these measurements to model results can provide insight on how that model's stratosphere impacts deduced emission footprints at various spatial and time scales. We find strong linear correlation between measured and model-derived methane DMFs across TCCON sites for both the tropospheric and total columns; however the zonal dependence of the agreement between model and measurements is not consistent for tropospheric versus total column methane. The model exhibits a strong interhemispheric difference in the contribution of stratospheric methane to the column, matching observations closely in the southern hemisphere and underestimating stratospheric methane in the northern hemisphere, which we attribute to dissimilarities in modeled partial column responsiveness to tropopause height. Discrepancies between vertical columns have a seasonal dependence, although biases increase over longer time periods at several sites. By comparing tropospheric partial columns of methane to those of GEOS-Chem, we can thus evaluate how the model's representation of seasonal and zonal fluctuations of methane mole fractions can introduce offsets in posterior emissions estimated by inversions.