With the availability of an increased number of ground-based weather radars, the development of composite radar rainfall estimates has become common practice. In mountainous terrain, weather radar measurements often encounter beam blockage effects, resulting in erroneous estimates of rainfall. This study introduces a novel relative radar quality index based on the radar reflectivity fraction to enhance the radar composite product. Additionally, we develop an improved mean field bias adjustment technique by including the spatial variability of the bias adjustment factors associated with the quality of radar observations. Radar reflectivity data from a network of single-polarization S-band radars, the Sattahip and Phimai radar stations in Thailand, and automatic rain gauges within the composite area, were used for the analysis. Three independent datasets were employed: (1) 51 storm events (2016–2022) for evaluating radar composite performance and QI-based bias adjustment; (2) hourly data from August–October 2020 to assess bias factor uncertainty; and (3) three heavy storms (2016, 2017, and 2020) to examine the QI method's effectiveness in beam-blocked basins. Our analysis explored seven combinations of hourly radar composite products. Subsequently, the performance of radar rainfall estimates obtained from applying the proposed mean field bias was evaluated by comparing them with the conventional technique. Results show the potential of integrating combined multiple quality indices to improve rainfall estimates, particularly for heavy rainfall events in mountainous regions.