Assessing 5 years of GOSAT Proxy XCH 4 data and associated uncertainties

摘要: Abstract. We present 5 years of GOSAT XCH4 retrieved using the "proxy" approach. The Proxy data are validated against ground-based TCCON observations and found to be high quality with a small bias 4.8 ppb (∼ 0.27 %) single-sounding precision 13.4 0.74 %). station-to-station (ameasure relative accuracy) is 4.2 ppb. For first time / XCO2 ratio component retrieval (bias 0.014 ppm−1 0.30 %), 0.033 0.72 %)). uncertainty relating model assessed through use an ensemble models. While each individual agree well validation (r = 0.94–0.97), it not possible select one as best from our comparisons. median value has smaller scatter (a standard deviation 0.92 ppm) than any models whilst maintaining (0.15 ppm). This used calculate maximum estimate uncertainty. compare this posteriori error (which assumed reduce sqrt(N)) find typically that becomes significant during summer months when at its lowest due increase in signal related increased summertime reflected sunlight. assess significance these uncertainties on flux inversion by comparing modelled TM5-4DVAR constrained NOAA surface (MACC reanalysis scenario S1-NOAA). for majority regions differences much larger estimated uncertainties. Our findings show useful information will provided inversions addition already assimilated measurements.