Toya ODAGIRI

Study on the Temporal Characteristics of Precipitation Based on the Particle Capture Characteristics of an Optical Reflection-Type Solid Precipitation Sensor

Tosiro KUMAKURA

This study aims to clarify how differences in the temporal aggregation scale affect precipitation estimation characteristics and error structures by using precipitation particle information observed with an optical reflection-type solid precipitation sensor (PDS: Precipitation Detection Sensor). Precipitation measurement provides fundamental data for understanding the hydrological cycle and is essential for flood prediction and disaster prevention planning. However, accurate observation of solid precipitation in winter is difficult because measurement errors can occur due to factors such as melting delay and reduced catch efficiency. Conventional rain gauges and disdrometers each have their own advantages and limitations. The PDS is an observational instrument designed to detect the passage of solid precipitation particles using reflected optical signals, enabling simultaneous estimation of precipitation amount and identification of precipitation type. In particular, the PDS is suitable for observing solid precipitation such as snowflakes and graupel and is expected to be a practical instrument for multi-site deployment.
In this study, precipitation amounts were estimated from the time-integrated particle mass derived from precipitation particle diameter and fall velocity observed by the PDS. The mass flux of particles was calculated from these particle properties, and precipitation types were classified using the CMF (Center of Mass Flux) method based on the relationship between particle diameter and fall velocity. The classified precipitation type information was then incorporated into the precipitation estimation procedure. In the analysis, multiple temporal aggregation intervals ranging from 5 minutes to 60 minutes were examined, extending beyond the conventional 5-minute interval commonly used in previous studies. The effects of different temporal aggregation scales on precipitation type composition, precipitation estimation results, and regression characteristics were compared. The observational data used in this study were obtained from PDS measurements at the Tokamachi experimental site in Niigata Prefecture. For comparison, observations from a Tamura-type precipitation intensity gauge and a Laser Precipitation Monitor (LPM) were also utilized.
The results show that as the temporal aggregation interval becomes longer, the correlation between the precipitation estimated by the PDS and the reference precipitation measurement improves, and error indices tend to decrease. In contrast, shorter aggregation intervals are more strongly affected by missing data and very light precipitation, resulting in larger variability and increased estimation errors. Furthermore, the composition ratio of precipitation types and the characteristics of particle distributions were found to vary depending on the aggregation interval, indicating that statistical variability in particle counts has a greater influence on estimation results at shorter temporal scales. These findings suggest that selecting an appropriate temporal aggregation interval according to the observation objective is essential for precipitation estimation using the PDS. This study systematically clarifies the temporal characteristics of PDS observations and provides fundamental knowledge that contributes to improving solid precipitation observations and precipitation estimation methods.