TROPOMI and WRF comparison for understanding South Sudan wetland emissions
MetadataShow full item record
Methane (CH4) is the second most important atmospheric greenhouse gas after CO2. Due to its shorter lifetime compared to CO2, it is a good target for short-term global warming mitigation. The recently launched TROPOspheric Monitoring Instrument (TROPOMI) provides us with atmospheric total column measurements of CH4 at an unprecedented combination of high spatial resolution of ~7x7 km2 and daily global coverage. Consequently, TROPOMI can allow us to detect and quantify localized sources of CH4 that can be reduced to effectively mitigate global warming. After only its first month in orbit, TROPOMI measured an unexpected high concentration of CH4 over the South Sudan region. In this study, we examine TROPOMI data in this region for the time period of November-December 2017, and we compare it with simulations of CH4 total column concentrations produced by the atmospheric transport model WRF-CHEM. We find that the model underestimates the CH4 spatial gradients mainly because of too low input emissions. We find that the model concentration enhancement, due to the wetland or anthropogenic emissions, is approximately 8 times lower than the enhancement measured by TROPOMI. Further, we analyze monthly averaged XCH4 data from TROPOMI. We find that monthly TROPOMI XCH4 enhancement follows the spatial pattern of high resolution wetland map and shows seasonality consistent with CH4 emissions from a process-based wetland model. Thus, it is likely that the enhancement is caused largely by CH4 emissions from wetlands. We also quantify CH4 emissions from TROPOMI data for November and December 2017 via a measurement-only (transport model-independent) approach. Overall, we estimate that the Sudd wetlands in the South Sudan region emitted 1057 ± 447 tonnes per hour CH4 in November 2017 and 566 ± 348 tonnes per hour CH4 in December, 2017. Comparing these results with the available CH4 emissions estimates from different process-based models, we find that the models highly underestimate the Sudd wetland emissions.