A Probabilistic Algorithm for Mitigating Persistent Ground Clutter in Doppler Weather Radar

The Doppler weather radars (DWRs) provide valuable three‐dimensional (3D) information about weather systems. The presence of objects such as trees, buildings, and mountains, also known as clutter, can significantly contaminate the reflectivity echoes. In complex terrains, signal processing techniques are often inadequate in completely resolving ground clutter. This study focuses on developing an algorithm to mitigate persistent ground clutter based on long‐term radar data. A composite ground clutter probability map was constructed using a texture‐based approach (Gabella filter), accessing persistent ground clutter across multiple scans. Otsu's thresholding method was then applied to determine an optimal threshold that separates clutter from non‐clutter regions. This results in a static binary clutter mask, which was further refined by morphological dilation. The performance of the proposed technique is evaluated on data from the C‐band DWR installed at the Thumba Equatorial Rocket Launching Station (TERLS) in Thiruvananthapuram, Kerala, India. The DWR observations from January and February 2017–2024 are used, characterized by the lowest average monthly rainfall over Kerala, maximizing clear‐air echoes. A total of 13,892 plan position indicator (PPI) scans (at 2‐degree elevation) were considered to generate the clutter map. The quantitative analysis of the clutter removal ratio indicates that the proposed technique effectively eliminates ground clutter, achieving of 0.98, compared to the standalone Gabella filter and fuzzy logic‐based methods for non‐rainy cases. This method offers a practical yet simple approach to mitigating clutter in complex terrains such as the Western Ghats (WG), as demonstrated in this study.