ISM Research Memorandum
No.
1080
Title:
Method for Evaluating the Equivalent and Total Error Associated with the Retrieval of CO2 Column Abundance from the GOSAT Satellite
Author(s):
Tomosada, Mitsuhiro (The Institute of Statistical Mathmatics);
Kanefuji, Koji (The Institute of Statistical Mathmatics);
Matsumoto, Yukio (Association of International Research Initiatives for Environmental Studies);
Tsubaki, Hiroe (The Institute of Statistical Mathmatics)
Key words:
Global warming, CO2 column abundance retrieval, satellite remote sensing, evaluation method.
Abstract:
Greenhouse Gases Observing Satellite (GOSAT) and Orbiting Carbon Observatory (OCO) for the purpose of observing greenhouse gases will be launched in early 2009 and in late 2008. GOSAT and OCO will be the first satellite to observe carbon dioxide (CO2), major greenhouse gases, from space. Global distributions of CO2 column abundances, retrieved from each GOSAT data and OCO data, will be opened to researchers and public users. So it is important to clarify the precision of the data retrieval. If satellite observations are to improve over the existing ground network, monthly averaged column data at a precision of 1% or better, for an 8~10 footprint are needed. The GOSAT mission was designed to observe CO2 density with 1% relative precision averaged in a certain period and with 10km~10km footprint size. The retrieval algorithm used in this study is followed by Rodgersfs method that is constrained Gauss-Newton method. Further, the evaluation method of the retrieved CO2 column abundance derived from the measurement error of spectrum is revealed by Rodgers.
In this paper, an evaluation method of the retrieved CO2 column abundance due to the errors of input parameters is represented when CO2 column abundance is retrieved. The input parameters, which are composed of the observation geometry and atmospheric conditions such as temperature profile, are needed. We expect to be obtained the observation geometry accurately. However, it is planned that atmospheric conditions at the observation time and at the observation point are obtained by interpolation of Global Analysis data (GANAL). Therefore, it is considered to be included some errors in the interpolated values. Then, we propose an evaluation method of the retrieved CO2 column abundance due to the errors of input parameters by the factorial experiments using an orthogonal array. And, evaluation method of the total error, which are derived from both the measurement error of spectrum and the errors of input parameters, are proposed. We apply the proposed evaluation method to the retrieved CO2 column abundance using simulated spectral data.