How can small radars be applied to addressing challenges in precipitation estimation by long-range large radars?

The S-band dual-polarization (large) radar, owing to its longer wavelength and lower susceptibility to attenuation, serves as the primary tool for quantitative estimation of large-scale precipitation. However, factors such as scanning methods, the Earth's curvature, and terrain blockage cause the beam center to be too high or even blocked during long-range detection, leading to a significant underestimation of precipitation in these areas. The deployment of X-band dual-polarization (small) radars offers a potential solution to this issue. Leveraging the advantage of X-band radar in low-altitude detection, this paper proposes a method to improve the data quality of S-band dual-polarization radar, aiming to address the underestimation problem in long-range precipitation estimation by large radars. Experiments were conducted based on six precipitation events observed by 9 X-band dual-polarization phased array radars in Guangdong and the Heyuan S-band dual-polarization radar. The scientific validity of the method was verified. The results show that in long-range detection areas, the algorithm can significantly mitigate the underestimation by large radar and achieve better precipitation estimation performance than either single-band (S or X) radar, reducing errors by at least 14.27%. For light to moderate rain, the primary improvement generated by the algorithm is a reduction in the mean error. For heavy to torrential rain, the algorithm significantly alleviates the severe underestimation by single-band radars, reducing estimation errors of heavy rain (hourly rainfall 30—50 mm) and torrential rain (hourly rainfall >50 mm) by at least 44.47% and 42.95%, respectively. The algorithm can accurately capture the rapid intensification signals of precipitation during heavy rainfall events, yielding results closely matching minute-level precipitation data observed by ground automatic stations. Compared to single-band radar, the algorithm demonstrates significant improvements in precipitation estimation, particularly in areas with dense X-band radar coverage. This method provides a valuable algorithmic reference for operational precipitation estimation using large and small radar network fusion in specific areas.