# Global stratospheric methane loss from satellite observations

## TL;DR

This study uses satellite data to show that methane loss in the stratosphere is higher than previously modeled, improving understanding of methane's role in climate and ozone chemistry.

## Contribution

The study provides the first observation-based estimate of stratospheric methane loss, revealing significant underestimation by models.

## Key findings

- Observational estimates of stratospheric methane loss are higher than model-based estimates.
- Incorporating observational data reduces the global methane budget imbalance from 23 to 3 Tg/y.
- Biases in methane concentrations and temperature drive discrepancies in model estimates.

## Abstract

Methane (CH4) is the second most important human-driven greenhouse gas after carbon dioxide, but its variability and long-term increase are not well understood, partly because of large uncertainties in how it is removed from the atmosphere. One important removal pathway is CH4 oxidation in the stratosphere, which also produces water vapor and reactive chemicals that affect climate and ozone. Until now, estimates of the stratospheric CH4 loss have relied only on models. Using satellite observations, we provide an observationally based estimate, showing that models systematically underestimate this loss. Incorporating our result into the global methane budget greatly reduces existing imbalances, helping reconcile top-down and bottom-up estimates and improving confidence in methane-climate assessments.

Stratospheric CH4 oxidation represents both an important sink in the global CH4 budget and a major source of stratospheric water vapor and hydrogen radicals, exerting strong influences on global climate and ozone chemistry. Yet, the magnitude of stratospheric CH4 chemical loss remains highly uncertain, with previous estimates largely relying on chemistry-climate models (CCMs). Here, we present an observationally based estimate of stratospheric CH4 loss (LSTR), derived from the CH4 diabatic flux across the isentropic surface fitted to the tropical tropopause, using satellite measurements of CH4 concentration, temperature, and radiative heating rates for 2007–2010. We obtain an LSTR of 49.8 ± 7.8 Tg/y, compared with 38.1 Tg/y estimated from reanalysis, and 25.7 Tg/y (range: 19.6 to 35.9 Tg/y) derived from CCMs, indicating that both reanalysis and CCMs systematically underestimate stratospheric CH4 loss. We show that discrepancies in global CH4 diabatic fluxes from the reanalysis and CCMs, when compared with observations, are mainly driven by biases in CH4 concentrations and further enhanced by errors in temperature and radiative heating. Substituting our observational estimate for the model-based stratospheric loss in the bottom-up global CH4 budget reduces the reported imbalance for the 2000s from 23 to 3 Tg/y, bringing it into close agreement with the 5 Tg/y (range: −4 to 13 Tg/y) imbalance inferred from top-down estimates. These findings highlight the critical role of observational constraints on LSTR in reconciling the global CH4 budget. They also carry important implications for understanding stratospheric water vapor and ozone chemistry.