Performance test of greenhouse gas retrieval algorithm over East Asia

In this study, current state of CO2 retrieval algorithm development is introduced and preliminary results are shown. This CO2 retrieval algorithm is based on optimal estimation method and utilized VLIDORT the vector discrete ordinate radiative transfer model to perform vector calculation. Proto type algorithm, developed from various combinations of state vectors to find accurate CO2 concentration, shows reasonable result on reference test. To generate synthetic GOSAT data several soundings which can represent East Asia atmosphere were simulated with high accuracy. Based on synthetic GOSAT data, reference test are performed to evaluate the performance of this algorithm and the errors induced by incorrect atmosphere information. According to results, aerosol properties (AOD, profile, type) are most important error inducing parameter. Retrieval error induced by a priori error of other atmospheric factors, wavenumber dispersion and geometrical error usually does not exceed 2 ppm. However errors in aerosol properties can induce more than 2ppm of retrieval error. This result shows the importance of aerosol information in CO2 retrieval. To improve CO2 retrieval the aerosol retrieval algorithm using GOSAT-CAI measurements, which provide aerosol information for the same area with GOSAT-FTS measurements, is utilized. The strategy for dealing with this information is also introduced.