Precipitation particles are fundamental elements constituting precipitation phenomena, and capturing their behavior is essential for understanding these phenomena. Individual particles exhibit diversity in particle size and fall velocity. Solid precipitation particles also have complex densities and shapes. Consequently, the understanding of cloud microphysical processes remains insufficient. Cloud microphysical processes are complex, consisting of precipitation particle collisions, water vapor diffusion, and spontaneous breakup. While drop size distribution (DSD) is obtained as a result of cloud microphysical processes, previous research has reported that DSD varies depending on latitude and regional characteristics. Therefore, this study aims to deepen the understanding of precipitation formation processes by focusing on the cloud microphysical processes of precipitation particles aloft in Japan, where both rain and snow are prevalent. In this study, a rainfall case during the rainy season in Shiga was analyzed, where vertical observations using VHF- and X-band radars were conducted. For the analysis of ice crystals and snowflakes, snowfall cases in Ishikawa and Hokkaido, which have different latitudinal zones and where Ka-band radar was installed, were extracted and analyzed using the satellite radar GPM/KuPR. Additionally, this study involves the development of analysis methods combining satellite and ground-based radars.

For raindrops, the estimation of DSD, which reflects cloud microphysical processes, was performed using vertical observation data from the X- and VHF-band radars. Combining these radars is advantageous for removing the influence of vertical air motion in the estimation of DSD. The estimated DSD assumed a gamma distribution with three unknown parameters representing intercept, slope, and shape parameters. To strictly evaluate the accuracy of the estimation method, this study analyzed stratiform precipitation where the effects of vertical air motion, turbulence within the radar observation volume, and non-uniform beam filling are considered small. A stratiform rainfall case on June 2, 2023, was selected. The slope and shape parameters of the DSD estimated from vertical observations tended to increase with decreasing altitude. This trend was similar to the vertical variation trend of parameters in tropical stratiform precipitation pointed out in previous studies. The vertical variation trends of the retrieved median volume diameter, liquid water content and normalized intercept parameter suggested that collision-coalescence was dominant in this case, followed by the contributions of breakup and evaporation. Previous studies dealing with tropical stratiform precipitation mentioned that the increasing trend of the shape parameter toward the lower layer was caused by evaporation. Thus, it was suggested that the cloud microphysical processes causing the vertical variation were different although the vertical variation trend of the raindrop size distribution parameters was the same in this case and the tropical case. The vertical variation of the median volume diameter and normalized intercept decreased toward the lower layer in this case, suggesting that collision-coalescence and breakup were approaching equilibrium states. In this case, it was confirmed from reanalysis data that the relative humidity at the analysis altitude was high. Consequently, the evaporation of small raindrops should be suppressed, and the collision opportunities between large raindrops and small raindrops increased, resulting in the predominance of collision-coalescence and collision-breakup.

For ice crystals and snowflakes, analyses were conducted using the Dual-Frequency Ratio (DFR), which theoretically reflects the median volume diameter. To obtain the DFR, analyses combining GPM/KuPR and ground-based Ka-band radar were conducted. Note that since satellite radar and ground-based radar are combined, the incident angle of the radar beam with respect to precipitation particles is different. When the DFR was calculated by combining GPM/KuPR and the ground-based Ka-band radar, larger DFRs were obtained in Ishikawa compared to Hokkaido for the same reflectivity. This suggests that ice crystals and snowflakes in Ishikawa have a larger average particle size than those in Hokkaido. The vertical distribution of temperature in Ishikawa was approximately 5 ℃ higher than that in Hokkaido, and the fact that the temperature near the ground was close to 0 ℃ in all cases in Ishikawa is considered to have worked advantageously for aggregation. Since the supersaturation with respect to ice at approximately -15 ℃ was higher in Ishikawa than in Hokkaido, aggregation by dendritic crystals was more promoted in Ishikawa. In addition, even in the low-temperature region from -15 to -20 ℃, which is lower than the temperature zone of the dendritic growth layer, larger DFRs were obtained in Ishikawa compared to Hokkaido for the same reflectivity. Since the supersaturation with respect to ice was larger in Ishikawa even in the lower temperature region, the growth by deposition of not only dendrites but also plate-like crystals should be larger in Ishikawa than in Hokkaido. This study shows that DFR obtained by combining satellite Ku-band and ground-based Ka-band radars provides a qualitative signature of the size of ice crystals and snowflakes.

This study developed methods to estimate DSD of raindrops and signatures related to the particle sizes of ice crystals and snowflakes through a comparison between two sites, utilizing various types of radars, including multi-frequency satellite and ground-based radars. This work will contribute to the advancement of accurate quantitative precipitation estimation in Japan.