Algorithm development for atmospheric gas remote sensing by measuring the high resolution spectrum and polarization

1.为了校正空间遥感大气气体中云和气溶胶对光的散射，已经发展了几种利用偏振的方法。假设用光谱分辨率为0.2cm^<-1>的傅立叶变换光谱仪在1.6 μ波段观测地球反射的太阳辐射，对气体的反演进行了模拟。结果表明，与不使用偏振光相比，使用偏振光可使CO_2色谱柱量的反演误差降低约0.5%。像移补偿（IMC）机制的不完全发挥会导致星载瞬态视场（IFOV）位置波动。利用卫星观测图像数据表明，这种波动会导致CO_2柱量的检索误差较大。为了纠正IFOV波动的影响，提出了一种利用FTS带通滤波器外的低频频谱来校正IFOV波动的方法。结果表明，应用该方法可大大减小误差。研究了地球公转、地球自旋运动和卫星运动对光谱线多普勒频移的影响。已经开发了一个程序来估计地外太阳光谱从地面高光谱分辨率的fts。利用地基高光谱分辨率FTS对CO_2进行了长期监测。初步精度试验表明，与塔台观测和空中观测结果吻合较好，CO_2相对值精度优于0.2%。

1.A few methods to use the polarization have been developed for correcting the scattering of light by cloud and aerosols in remote sensing atmospheric gases from space. Simulations of retrieval of gases have been performed assuming the observation of Earth-reflected solar radiation is made in the 1.6 μ band with a FTS (Fourier Transform Spectrometer) having a spectral resolution 0.2cm^<-1>. It has been shown that the retrieval error of CO_2 column amount is reduced by about 0.5 % when the polarization was used compared with non-use of polarization.2.Incomplete performance of IMC (Image Motion Compensation) mechanism generates a fluctuation of the position of IFOV (Instantaneous Field of View) of satellite borne FTS. It has been shown, by using satellite-observed image data, that this fluctuation of IFOV generates large retrieval error of CO_2 column amount. To correct the effect of IFOV fluctuation, a method has been developed by using the low frequency spectrum out side the band pass filter of FTS. It has been found that the error is greatly reduced by applying this method.3.The effect of Doppler shift of spectral lines due to the Earth revolution around the sun, spin motion of the Earth and satellite motion has been investigated.4.A program has been developed to estimate the extra-terrestrial solar spectrum from ground-based high spectral-resolution FTS.5.A long term monitor of CO_2 amount from ground-based high spectral-resolution FTS has been initiated. The preliminary test of accuracy showed a good coincidence with the tower observation and air borne observation, showing the accuracy of relative value of CO_2 better than 0.2 %.