Bias corrections of GOSAT SWIR XCO2 and XCH4 with TCCON data andtheir evaluation using aircraft measurement data

摘要

We describe a method for removing systematic biases of column-averaged dryair mole fractions of CO (XCO) and CH (XCH) derivedfrom short-wavelength infrared (SWIR) spectra of the Greenhouse gasesObserving SATellite (GOSAT). We conduct correlation analyses between theGOSAT biases and simultaneously retrieved auxiliary parameters. We use thesecorrelations to bias correct the GOSAT data, removing these spuriouscorrelations. Data from the Total Carbon Column Observing Network (TCCON) wereused as reference values for this regression analysis. To evaluate theeffectiveness of this correction method, the uncorrected/corrected GOSAT datawere compared to independent XCO and XCH data derived fromaircraft measurements taken for the Comprehensive Observation Network forTRace gases by AIrLiner (CONTRAIL) project, the National Oceanic andAtmospheric Administration (NOAA), the US Department of Energy (DOE), theNational Institute for Environmental Studies (NIES), the Japan MeteorologicalAgency (JMA), the HIAPER Pole-to-Pole observations (HIPPO) program, and theGOSAT validation aircraft observation campaign over Japan. These comparisonsdemonstrate that the empirically derived bias correction improves theagreement between GOSAT XCO/XCH and the aircraft data. Finally,we present spatial distributions and temporal variations of the derived GOSATbiases.