A Land–Sea Contamination Suppression Method for Interferometric Microwave Radiometer Based on Nodal Sampling

Interferometric microwave radiometers often suffer from land-sea contamination (LSC) in coastal brightness temperature (TB) observations due to the Gibbs effect. This contamination leads to anomalously high nearshore TB, significantly affecting data quality and subsequent applications. The nodal sampling (NS) can correct this contamination but have limitations under complex coastal conditions. This study proposes an improved NS algorithm targeted at the coastal regions, named LSC-NS. The method enhances nearshore details through spatial oversampling and constructs sinusoidal surrogate signals in contaminated regions. The method was first validated using an interferometric microwave radiometer (IMR) simulation system. Then, SMOS data were used for three-level evaluation. In orbit validation over open sea regions, the average normalized root mean square error (NRMSE) decreased from 14.97% to 7.91%. In more challenging closed sea regions, the average NRMSE dropped from 33.52% to 25.05%, with a significant reduction in coefficient of variation as well. These results demonstrate that the proposed algorithm effectively mitigates land contamination and substantially improves TB accuracy and spatial consistency. The method provides essential algorithm support for operational applications of IMRs in coastal areas and offers valuable reference for similar future satellite missions.