<![CDATA[Ground clutter remains a significant source of contamination in weather radar observations, adversely affecting the interpretation of echoes and subsequent meteorological applications. This study assesses the Dual-Polarisation Dual-Scan (DPDS) framework. This Bayesian-based classifier combines polarimetric descriptors (ρₕᵥ, ZDR) with temporal coherence (ρ₁₂) derived from consecutive azimuthal scans. The analysis uses I/Q data from an operational dual-polarisation C-band weather radar located in Sidoarjo, Surabaya, Indonesia. Results indicate that the DPDS framework significantly outperforms traditional Dual-Polarisation (DP) and Dual-Scan (DS) methods. For moving weather (W), the DPDS achieved a Probability of Detection (POD) of 0.939, a 313-fold improvement over the DP-only method, which suffered from severe polarimetric overlap between clutter and rain. While the clutter class exhibited a False Alarm Ratio (FAR) of 0.749, this is attributed to the 83-second scan interval of the Sidoarjo radar; over this duration, stable tropical rain remains highly correlated, mimicking the temporal signature of stationary ground clutter (C). However, the framework successfully preserved the integrity of the meteorological field, reducing the misclassification of zero-velocity weather (W0) compared to DS-only methods and achieving an overall accuracy of 0.982. These findings highlight the effectiveness of integrating polarimetric and temporal decorrelation information to establish a more robust, physically consistent echo classification framework, particularly under challenging conditions of clutter and low-velocity weather.]]>
