| The geostationary satellite television (TV) signals that are reflected off the ocean surfaces could enter the AMSR-E antenna, resulting in RFI (Radio Frequency Interference) contamination in AMSR-E 10.65 and 18.7 GHz channels. If not detected, the presence of RFI signals can result in false retrievals of oceanic environmental parameters (e.g., sea surface temperature, sea surface wind speed, rain water path) from microwave imaging radiance measurements. This study first examined the geometric relationship between the RFI source, geostationary TV satellite, and AMSR-E observation. Then a normalized Principal Component Analysis (NPCA) method is proposed and applied for RFI detection over oceans in Advanced Microwave Scanning Radiometer (AMSR)-E observations. It is found that the RFI-contaminated observations on AMSR-E descending node at 10.65 and 18.7 GHz can be successively detected near coastal areas surrounding Europe and United States continents. The results yielded from the geometric examination at another angle verify those signals detected with NPCA. The proposed NPCA algorithm is applicable in an operational environment for fast data processing and data dissemination, and is different from earlier methods, which often require a priori information. |
