Study On Retrieval Algorithm Of Cloud Liquid Based On Data Measured By Airborne Microwave Radiometer With Two Antennas

Cloud liquid water content is always one of the most important quantities in many subjects and operations, particularly the conditions of cloud liquid water and supercooled liquid water content are the predetermination condition of cloud seeding in weather modification. An airborne up-looking microwave radiometer with single antenna was invented by the institute of Atmospheric Physics, Chinese Academy of Sciences cooperated with Jilin weather modification office and the institute of Geograpy, Changchun in 2001. The path-integrated cloud liquid water content can be determined by measuring of microwave emission from the cloud with the instrument and inversing the obtained data with statistical method. Now the instrument needs updating. The single antenna is replaced by two fixed antennas that elevated equal angle of 45 degree above the longitudinal axis of the aircraft with one pointing forward, the other pointing afterward. A complete vertical cross section of the distribution of cloud water can be obtained by using the new instrument with topographic method. The method involves measurement of microwave emission from cloud in many locations under a cloud and mathematical inversion of this data to obtain the liquid water distribution. The mathematical inversions of the data include two steps: the first step is to get path-integrated cloud liquid water content of each measurement with, the statistical method that has been used before. In the second step the retrieval domain is divided into square elements and the linear equations of the relationship that between path-integrated cloud liquid water content of each measurement and parts of it are made, then the distribution of liquid water through a cross section of a cloud can be calculated with LSQR algorithm.As main part of the method, the second step is presented in this paper and applied to simulated measurements to determine the accuracy attainable with different kinds of cloud models and under various conditions. The results of the simulation study show that the technique is capable of yielding the liquid water content with a spatial resolution of a few hundred meters;the retrieval accuracy determined by the size of the cloud, the maximum water content, the resolution, andthe number of measurements. A single center shaped objective cloud with dimension of 7000*7000 and maximum liquid content of 1.2g/m3 is measured and divided with grids of 583 meter resolution. Results shows that the retrieval maximum relative error is 11.9% and trie root mean square error is 0.014g/m3 with 199 times of microwave radiometer measurement and the maximum noise of 5.0g/m2.