Tomographic Retrieval Of Atmosphere Temperature From Limb Infrared Emission

With the large amount of greenhouse gas emission and big deforestation, the greenhouse effect causes the atmospheric temperature to be a higher level and more extreme climate happen. People pay a more attention to the increasing atmospheric temperature, and more and more governments work together to limite the greenhouse emission. Through long-term global and accurate atmospheric system observation, people can understand and master the changes of atmosphere and environment, then can provide routine meteorological factors prediction, and other physical parameters of the quantitative prediction of the atmosphere, and also provide the countermeasures proposed for the survival and development of human. It is becoming more and more important to observe the Atmospheric parameters and its contents.Atmospheric Limb observation mode, in which the light of sight is parallel to the surface of the Earth, has the advantages of Occultation observation mode and Nadir model, which can be both a global coverage and fine vertical resolution of the atmosphere. The MIPAS on European Space Agency's ENVISAT works in this type modle. Right now, Observation and research in Limb model carried out by my country is still very little. On the national "863" project support, the fundamental principles of the Limb observation model, as well as the atmospheric state parameters simulation (such as temperature inversion) is studingIn this paper, based on atmospheric sounding instrument MIPAS, the temperature of one-dimensional atmospheric model and two-dimensional atmospheric models, are inversed. In the process, to solve the infrared transfer equation which is ill-condition equation, this should be regularized, and Newton nonlinear iteration is used to get the atmospheric temperature inversion.In the one-dimensional atmospheric model, one Limb sequence MIPAS L1 data was selected to reverse the temperature profile. First we chose the Microwindow that L2 product used to reduse the redundance data. Then we compare the result with the MIPAS L2 product that use the same L1 data of MIPAS. The differences between the reversed result and the product L2 temperature is less than 0.7K.The differences may be due to the different transfer model used; After that, based on the medical field computed tomography principle as the starting point, by combining tomography with atmospheric radiative transfer theory, and on the two-dimensional atmospheric field, the numerical inversion of atmospheric simulation was performed. In the numerical inversion simulations of two-dimensional atmospheric model, the basic parameters of the MIPAS sensor, spectral resolution reduced from the nominal 0.025 cm-1 to 0.1 cm-1, thereby the time for a single observation sequence Limb reduce to about original quarter, and at the same time this increases the number of Limb sequence 4th times greater, thus increased Limb sequence overlap, making it more consistent with the multi-angle tomography observation conditions.In the two-dimensional atmospheric field, firstly, the 2D reference temperature field was made with the data extracted from the MPI Mainz database from north latitude 85°to north latitude 5°of July. The simulation of two-dimensional atmospheric radiation and one-dimensional atmospheric models were compared, the two radiational differences between the value of simulation in some bande even reachup to ten times bigger than the noise equivalent, and then by the perturbation analysis, this is due to different sensitivity to the different positions for different bands.In the Two-dimensional atmospheric temperature field, using the selected microwindow for two-dimensional atmospheric model, a one-dimensional inversion method was performed.It makes the result reach up to a large difference about 14K. With the tomography method the difference is less than 2K and the standard deviation is 0.54K. Tomography method also makes the results of inversion more better horizontal resolution and vertical resolution opposed to the conventional model. Compareed with the normal Limb observations, the 2D model have greatly improved the inversion. This is due to the observation by multi-angles, and the usefull extraction of multi-angle observations of the information, on the other hand, two-dimensional atmospheric model can use the spectral band that the one-dimensional atmospheric avoid to use because the atmosphere is not homogeneous ,thereby it may also increase atmospheric inversion accuracy.