An increase in methane emissions from tropical Africa between 2010 and 2016 inferred from satellite data

Abstract

Abstract. Emissions of methane (CH4) from tropical ecosystems, and how they respond to changes in climate, represent one of the biggest uncertainties associated with the global CH4 budget. Historically, this has been due to the dearth of pan-tropical in situ measurements, which is particularly acute in Africa. By virtue of their superior spatial coverage, satellite observations of atmospheric CH4 columns can help to narrow down some of the uncertainties in the tropical CH4 emission budget. We use proxy column retrievals of atmospheric CH4 (XCH4) from the Japanese Greenhouse gases Observing Satellite (GOSAT) and the nested version of the GEOS-Chem atmospheric chemistry and transport model (0.5∘×0.625∘) to infer emissions from tropical Africa between 2010 and 2016. Proxy retrievals of XCH4 are less sensitive to scattering due to clouds and aerosol than full physics retrievals, but the method assumes that the global distribution of carbon dioxide (CO2) is known. We explore the sensitivity of inferred a posteriori emissions to this source of systematic error by using two different XCH4 data products that are determined using different model CO2 fields. We infer monthly emissions from GOSAT XCH4 data using a hierarchical Bayesian framework, allowing us to report seasonal cycles and trends in annual mean values. We find mean tropical African emissions between 2010 and 2016 range from 76 (74–78) to 80 (78–82) Tg yr−1, depending on the proxy XCH4 data used, with larger differences in Northern Hemisphere Africa than Southern Hemisphere Africa. We find a robust positive linear trend in tropical African CH4 emissions for our 7-year study period, with values of 1.5 (1.1–1.9) Tg yr−1 or 2.1 (1.7–2.5) Tg yr−1, depending on the CO2 data product used in the proxy retrieval. This linear emissions trend accounts for around a third of the global emissions growth rate during this period. A substantial portion of this increase is due to a short-term increase in emissions of 3 Tg yr−1 between 2011 and 2015 from the Sudd in South Sudan. Using satellite land surface temperature anomalies and altimetry data, we find this increase in CH4 emissions is consistent with an increase in wetland extent due to increased inflow from the White Nile, although the data indicate that the Sudd was anomalously dry at the start of our inversion period. We find a strong seasonality in emissions across Northern Hemisphere Africa, with the timing of the seasonal emissions peak coincident with the seasonal peak in ground water storage. In contrast, we find that a posteriori CH4 emissions from the wetland area of the Congo Basin are approximately constant throughout the year, consistent with less temporal variability in wetland extent, and significantly smaller than a priori estimates.

Attribution of the accelerating increase in atmospheric methane during 2010–2018 by inverse analysis of GOSAT observations

“…The downregulation of North American emissions is consistent with recent WetCHARTs updates (v1.2.1) represent substantially reduced methane emissions across regions categorized as partial wetland complexes (Lehner & Döll, 2004;Bloom et al, 2017). Recent studies found that WetCHARTs overestimates wetland emissions in the Congo Basin but underestimates in the Sudd region (Lunt et al, 2019;Parker et al, 2020b;Pandey et al, 2020). Our inversion is unable to resolve this spatial correction pattern, because of coarse resolution in the wetland state vector (both regions are in Sub-Sahara Africa, i.e., wetland region 10 in Figure 1).…”

“…A number of inverse analyses previously used GOSAT observations to constrain methane emission estimates (Fraser et al, 2013;Monteil et al, 2013;Cressot et al, 2014;Alexe et al, 2015;Turner et al, 2015;Pandey et al, 2016;Pandey et al, 2017a;Miller et al, 2019;F. Wang et al, 2019a;Lunt et al, 2019;Maasakkers et al, 2019;Janardanan et al, 2020;Tunnicliffe et al, 2020;Yin et al, 2020). Maasakkers et al (2019) used 2010-2015 GOSAT observations to optimize gridded methane emissions, global OH concentrations, and their 2010-2015 trends.…”

Country-Scale Analysis of Methane Emissions with a High-Resolution Inverse Model Using GOSAT and Surface Observations

“…The inverse model suggested an overall underestimation in the prior for equatorial African countries (Figure 4f), such as Uganda, Tanzania, Sudan, and Kenya, though the annual emissions were lower for these countries. A recent study ( [60]) using GOSAT XCH 4 observations in their inversion reported overall larger emissions compared to prior over Africa with strong exceptions in the Congo basin. However, in our analysis, we found a slight increase in our posterior emissions over the Democratic Republic of Congo.…”

“…Tropical Africa is also a natural methane emitter (12% of global wetland emission, [59]) where the sources are wetlands, flood plain, riverine ecosystems, etc. Due to the seasonal migration of the intertropical convergence zone (ITCZ), the inundation extent is highly variable in these water bodies, and thus there is significant variability in the estimates of methane emission in this region ( [60]) and difficulty in models to capture the wetland emissions. Significant reductions were observed over boreal North America and Russia (Figure 3).…”

Global methane budget and trend, 2010–2017: complementarity of inverse analyses using in situ (GLOBALVIEWplus CH<sub>4</sub> ObsPack) and satellite (GOSAT) observations

“…Inverse analyses of atmospheric methane observations using chemical transport models (CTM) provide a formal method for inferring methane emissions and their trends (Brasseur and Jacob, 2017). Global satellite observations of atmospheric methane columns from the shortwave infrared SCIAMACHY and GOSAT instruments have been widely used for this purpose (Bergamaschi et al, 2013;Wecht et al, 2014;Turner et al, 2015;Maasakkers et al, 2019;Miller et al, 2019;Lunt et al, 2019). Other inverse analyses have relied on in situ methane observations that have much higher precision, are more sensitive to surface emissions, and include isotopic information, but are much sparser (Pison et al, 2009;Bousquet et al, 2011;Miller et al, 2013;Patra et al, 2016;McNorton et al, 2018).…”