大气遥感辐射传输模式的一种有效的Jacobian算法

介绍并验证了一种快速、精确地计算Jacobian的解析方法(辐射传输方程的线性化,简称LR方法),将其结果与NAST-I真正的线性模式及数值方法(差分方法)的Jacobian值进行比较。结果表明,这种解析算法不仅快速、精确,而且独立于辐射传输方程,在有效Jacobian计算中只需要输入大气透过率,可适用于任何垂直探测器权重函数的计算。

A Efficient Jacobian Algorithm for Radiative Transfer Models of Atmospheric Remote Sensing

Using the high spectral resolution infrared sounder radiance measurements,fast and accurate calculations for radiative transfer model(RTM) and its Jacobian is needed in order to inverse the radiances or brightness temperature(BT) for atmospheric profiles or assimilate the radiances in Numberical Weather Prediction(NWP) in near real time.Linearization of radiative transfer equation(RTE)is the key in physical retrieval of atmospheric temperature and moisture profiles from infrared sounder observations.In this paper,one fast and accurate method is described for computing the Jacobians,namely from linearized RTE(LR method).The methods are independent of RTMs,only atmospheric transmittances are required in the efficient Jacobian calculations.The NAST-I OSS(NAST-I: National Polar-Orbiting Operational Environmental Satellite System Airborne Sounder Testbed-Interferometer Instrument,OSS: Optimal Spectral Sampling)RTM model and numerical method are used for testing the computations.The derived Jacobians are compared with the linear model that OSS RTMs and numerical method provided.The results show that Jacobians from the fast computations are very close to that from the true linear model,while the fast methods takes only 0.5 TB calculations.