The effects of rain on a Ka-band swath altimeter: lessons learned from the SWOT mission

The Surface Water and Ocean Topography (SWOT) mission offers unprecedented Ka-band swath altimetry measurements via its KaRIn instrument, but remains highly sensitive to signal attenuation by precipitation. This study investigates the radiometric behavior of KaRIn under rain conditions, focusing on the characterization, correction, and physical interpretation of the normalized radar backscatter coefficient (sigma0). A three-regime decibel conversion scheme was implemented to handle linear sigma0 values, including negative returns, and a parametric angular correction model was applied based on wind-dependent polynomial fits. Cross-validation against KaPR (GPM) and AltiKa revealed consistent angular trends and wind dependencies, with systematic biases of +2.3 dB and +3.3 dB, respectively, over wind speeds ranging from 3 to 13 m/s, which account for over 85% of oceanic conditions globally. Two rainfall retrieval methods were developed from KaRIn sigma0: a physically-based attenuation inversion using the ITU-R gamma–R relation, and a supervised random forest (RF) classifier trained with collocated NEXRAD ground radar measurements. The RF model achieved an overall accuracy of 89.2%, with a detection probability of 82.5% for rain rates above 5 mm/hr, compared to 72.4% for the ITU approach. Global analysis confirms that rain rates exceeding 5 mm/hr or an attenuation of 10 dB result in significant degradation of KaRIn sea surface height (SSH) retrievals. Above this threshold, more than 95% of SSH observations are rejected by Level-3 editing filters, validating the statistical relevance of the rain flag criterion. Beyond SWOT, this study provides a methodological foundation for Ka-band altimetry in upcoming missions. The Sentinel-3 Next Generation (S3-NG) mission will benefit from these rain detection algorithms during post-launch calibration and data quality control. Similarly, the ODYSEA mission—a CNES–NASA Doppler scatterometer designed to resolve fine-scale vector winds and surface currents—will rely on accurate rain filtering to isolate geophysical signals. The statistical characterization of Ka-band attenuation and the rain retrieval strategies presented here are key to enabling reliable Ka-band remote sensing in dynamic meteorological environments.