Validation of XCH<sub>4</sub> derived from SWIR spectra of GOSAT TANSO-FTS with aircraft measurement data

Abstract

Abstract. Column-averaged dry-air mole fractions of methane (XCH 4 ), retrieved from Greenhouse gases Observing SATellite (GOSAT) short-wavelength infrared (SWIR) spectra, were validated by using aircraft measurement data from the National Oceanic and Atmospheric Administration (NOAA), the US Department of Energy (DOE), the National Institute for Environmental Studies (NIES), the HIAPER Pole-to-Pole Observations (HIPPO) program, and the GOSAT validation aircraft observation campaign over Japan. In the calculation of XCH 4 from aircraft measurements (aircraft-based XCH 4 ), other satellite data were used for the CH 4 profiles above the tropopause. We proposed a data-screening scheme for aircraft-based XCH 4 for reliable validation of GOSAT XCH 4 . Further, we examined the impact of GOSAT SWIR column averaging kernels (CAK) on the aircraft-based XCH 4 calculation and found that the difference between aircraft-based XCH 4 with and without the application of the GOSAT CAK was less than ±9 ppb at maximum, with an average difference of −0.5 ppb.We compared GOSAT XCH 4 Ver. 02.00 data retrieved within ±2 • or ±5 • latitude-longitude boxes centered at each aircraft measurement site with aircraft-based XCH 4 measured on a GOSAT overpass day. In general, GOSAT XCH 4 was in good agreement with aircraft-based XCH 4 . However, over land, the GOSAT data showed a positive bias of 1.5 ppb (2.0 ppb) with a standard deviation of 14.9 ppb (16.0 ppb) within the ±2 • (±5 • ) boxes, and over ocean, the average bias was 4.1 ppb (6.5 ppb) with a standard deviation of 9.4 ppb (8.8 ppb) within the ±2 • (±5 • ) boxes. In addition, we obtained similar results when we used an aircraft-based XCH 4 time series obtained by curve fitting with temporal interpolation for comparison with GOSAT data.

Bias corrections of GOSAT SWIR XCO<sub>2</sub> and XCH<sub>4</sub> with TCCON data and their evaluation using aircraft measurement data

“…Additionally, we take the mean difference between TCCON XCH 4 and aircraft-based XCH 4 into account when calculating the coefficients C 0 over land and ocean for XCH 4 . In a comparative analysis at four locations, Inoue et al (2014) showed that aircraft-based XCH 4 was 8.6 ppb smaller than TCCON XCH 4 on average. Consequently, in this study, we used the values for C 0 shown in Table 3b (i.e., 6.0 ppb over land and 0.2 ppb over ocean) for the correction of GOSAT XCH 4 when comparing GOSAT XCH 4 with TCCON XCH 4 .…”

“…To calculate aircraft-based XCO 2 and XCH 4 (as described in the next paragraph), we also used tower data from the Meteorological Research Institute (MRI) in Tsukuba (Inoue andMatsueda, 1996, 2001) and the NOAA ESRL/GMD tall tower network in Park Falls, WI and West Branch, IA . Details of the aircraft and tower measurements are described in Inoue et al (2013) and Inoue et al (2014), except for the JMA aircraft and groundbased measurements. The JMA aircraft measurements are conducted by utilizing the cargo aircraft C-130H of the Japan Ministry of Defense (MOD) to collect flask air samples once a month during a regular flight between the mainland of Japan and Minamitorishima, an island located nearly 2000 km southeast of Tokyo (Tsuboi et al, 2013).…”

“…Figure 1b and Table 2 show a global map and basic information, respectively, on every aircraft measurement site used in this study. Aircraft-based XCO 2 and XCH 4 are calculated by applying the GOSAT SWIR column averaging kernels (CAK) by using the methods developed by Miyamoto et al (2013) and Inoue et al (2014), respectively. There is one difference in the aircraft XCH 4 calculation: Inoue et al (2014) used fixed monthly climatologies for the CH 4 profiles above the tropopause and did not include the yearly trend in CH 4 concentration because of their short analysis period (June 2009to July 2010.…”

“…02.00) are biased −1-2 ppm (±1-3 ppm) against the aircraft measurement data , whereas GOSAT SWIR XCH 4 measurements (Ver. 02.00) are biased positively by 2-7 ppb with a standard deviation of about 15 ppb (Inoue et al, 2014).…”

“…Column-averaged dry air mole fractions of CO 2 (XCO 2 ) and CH 4 (XCH 4 ) are retrieved from the Short-Wavelength InfraRed (SWIR) spectra of the Thermal And Near-infrared Sensor for carbon Observation -Fourier Transform Spectrometer (TANSO-FTS) onboard GOSAT. Validation of XCO 2 and XCH 4 derived from the GOSAT TANSO-FTS has been conducted by using ground-based high-resolution Fourier Transform Spectrometer (ground-based FTS) data and aircraft measurements Saitoh et al, 2012;Yoshida et al, 2013;Inoue et al, 2013Inoue et al, , 2014Gavrilov et al, 2014). The results showed that the GOSAT SWIR XCO 2 measurements (Ver.…”