A41H-3153:
Correction of Spectral Distortion on Oxygen a-Band Due to Non-Linear Phase Delay Onboard Tanso-FTS/Gosat

Thursday, 18 December 2014
Hiroshi Suto1, Andre Butz2 and Akihiko Kuze1, (1)JAXA Japan Aerospace Exploration Agency, Sagamihara, Japan, (2)Karlsruhe Institute of Technology, Karlsruhe, Germany
Abstract:
To observe the global column concentration of carbon dioxide (CO2) and methane (CH4) from space, the Greenhouse gases Observing SATellite (GOSAT) was launched on January 23, 2009, and has started the operational observation. Thermal and Near Infrared Sensor for Carbon Observation – Fourier Transform Spectrometer (TANSO-FTS) has been continuously measuring CO2 and CH4 distributions globally, and the retrieved column CO2 and CH4 data have been distributed to the public. To make a successful retrieval of XCO2 and XCH4, the spectral quality of Oxygen A-band is the most importance. Over five years in-orbit operation of TANSO-FTS, the spectral distortion related with input radiance on Oxygen A-band have been observed and reduced the retrieval accuracy and precision of XCO2 and XCH4. It suggests that the Oxygen A-band signal chain has non-linear response against input radiance. To characterize the non-linear response of signal chain against input signal levels, the test procedure is newly developed coupled with the modulated laser light, simultaneous signal acquisition system and on-ground TANSO-FTS, which called engineering model. The results present clearly that the analogue signal chain of Oxygen A-band excites the non-linear response both of amplitude and phase delay against input signal levels. Also, the non-linear interferogram drives both of the artificial spectra on the out-band region and the spectral distortion linked with absorption spectral lines. To improve the spectral quality of Oxygen A-band, these artificial and distorted spectra have to correct with properly. The newly correction algorithm for level-1 processing was developed and the corrected spectra were retrieved and validated by applying RemoTeC algorithm. Comparing with the previous version of level-1 products, the agreement between observation and theoretical calculation is well improved and the biases of biases of XCO2 and XCH4 against ground validation site are reduced.