The imprint of stratospheric transport on column-averaged methane

Abstract. Model simulations of column-averaged methane mixing ratios (XCH₄) are extensively used for inverse estimates of methane (CH₄) emissions from atmospheric measurements. Our study shows that virtually all chemical transport models (CTM) used for this purpose are affected by stratospheric model-transport errors. We quantify the impact of such model transport errors on the simulation of stratospheric CH₄ concentrations via an a posteriori correction method. This approach compares measurements of the mean age of air with modeled age and expresses the difference in terms of a correction to modeled stratospheric CH₄ mixing ratios. We find age differences up to ~ 3 years yield to a bias in simulated CH₄ of up to 250 parts per billion (ppb). Comparisons between model simulations and ground-based XCH₄ observations from the Total Carbon Column Network (TCCON) reveal that stratospheric model-transport errors cause biases in XCH₄ of ~ 20 ppb in the midlatitudes and ~ 27 ppb in the arctic region. Improved overall as well as seasonal model-observation agreement in XCH₄ suggests that the proposed, age-of-air-based stratospheric correction is reasonable.

The latitudinal model bias in XCH₄ is supposed to reduce the accuracy of inverse estimates using satellite-derived XCH₄ data. Therefore, we provide an estimate of the impact of stratospheric model-transport errors in terms of CH₄ flux errors. Using a one-box approximation, we show that average model errors in stratospheric transport correspond to an overestimation of CH₄ emissions by ~ 40 % (~ 7 Tg yr⁻¹) for the arctic, ~ 5 % (~ 7 Tg yr⁻¹) for the northern, and ~ 60 % (~ 7 Tg yr⁻¹) for the southern hemispheric mid-latitude region. We conclude that an improved modeling of stratospheric transport is highly desirable for the joint use with atmospheric XCH₄ observations in atmospheric inversions.