Detection Of Radio-frequency Interference From Microwave Radiometer Data And Its Influence On Surface Soil Moisture Retrieval

With the improvement of weather forecasting and the development of technology, the measurements of meteorological satellite have gradually become an important supplement to the conventional observations. Among the satellite sensors, microwave radiometers can provide observation data day and night, it has good penetration can be not effected by cloud. The surface soil moisture retrieval is one of the most important applications of observation data from satellite-based passive microwave remote sensing. But the RFI signals from brightness temperatures will pollute the observations and reduce its quality and application. So it is very important to detect the RFI from brightness temperatures and correct RFI contaminated data before the data be used in retrieve or assimilation.In this study, the double principal component analysis (DPCA) method is used to identify RFI signals from microwave sensing observations and tested its affection on different measurements of microwave remote sensing and different land surface conditions. Then, use this method to get a long time RFI signals and analysis its variation trend with time. Also, by using the original brightness temperatures data and the data after RFI detection and correction to retrieve soil moisture, the influence of RFI on retrieval results is researched.The main contents and conclusions in this thesis are summarized as following.(1) Radio Frequency Interference (RFI) causes severe contamination to the passive and active microwave sensing observations and corresponding retrieval products over some continents. RFI signals should be detected and filtered before applying the microwave data to retrieval and data assimilation. It is difficult to detect RFI over land surfaces covered by snow because of the scattering effect of snow surface. The double principal component analysis (DPCA) method is adopted in this study, and its ability in identifying RFI signals in AMSR-E data over snow covered regions is investigated. Results show that DPCA method can detect RFI signals effectively in spite of the impact of snow scattering. And the detected RFI signals persistent with time. Compared to other methods, such as PCA and normalized PCA, DPCA is more robust and suitable for operational application.(2) To further verify the applicability and effectiveness of DCPA method, it has been used to detect the RFI signals from brightness temperatures data of MWRI on-aboard FY-3B. The results show that this method can identify the RFI signals fro MWRI data over land and ocean. Over land, the RFI signals at 10.65GHz distributes widely over Europe and Japan, and less over the United States and China. Over ocean, the RFI often occur in the coastal area. AMSR-E observations of 18.7 GHz are contaminated by RFI along the coastline of America, and 10.65 GHz often occur in Mediterranean.(3) For the application of long time passive microwave observations to climate research, use the DPCA method to identify the RFI signals from AMSR-E brightness temperatures during 2002 September to 2011 September and get the variation of RFI with time. From the result it can be found that the distribution and intensity of RFI signals will change with time. And the number of RFI signals on C-band data decreases with time, and X-band increases.(4) The DCPA method is used to detect the RFI signals from AMSR-E data and its influence on the retrieval of soil moisture is studied. Then, the data polluted by RFI is corrected by linear fitting. Comparing the retrieved products by the land parameter retrieval model (LPRM) method before and after RFI correction, it can been found that the soil moisture interfered by RFI is not convergence. So it is very necessary to effectively identify and correct RFI prior to observations with space-borne microwave imagers to retrieve or assimilation.