2026 年 82 巻 3 号 p. 138-150
Winter nocturnal temperature drops in cold regions arise from complex interactions between radiative cooling and terrain-induced micrometeorological processes. Using long-term operational surface observations from 1978 to 2025, this study investigated the regional and topographic controls on nocturnal cooling in Hokkaido, Japan. The results revealed a pronounced regional contrast in long-term mean temperature drops, with stronger cooling on the Pacific side and weaker cooling on the Sea of Japan side. These differences reflect variations in nocturnal radiative conditions, inferred indirectly from the sunshine duration during the 24-h noon-to-noon period, and are further modulated by local cold-air pooling characteristics, as represented by clear-night wind speed. With respect to extreme temperature drops, principal component analysis identified southeastern and northern Hokkaido as the regions most susceptible to intense and widespread nocturnal cooling. A key contribution of this study is the introduction of two terrain-based metrics that capture the terrain characteristics responsible for such large temperature drops, extending the analysis beyond the basin- and valley-focused approaches common in previous studies. These metrics are Rav, the area-volume ratio representing the radiative-cooling capacity of the surrounding terrain, and Asite/Ain, the fraction of the inner-catchment area lying below the site elevation, representing terrain openness and the site’s relative elevation. Analysis indicated that cooling in shallow terrain is governed primarily by Rav, whereas cooling in deep terrain depends on the cold-air pooling structure and the site’s relative position within the pool, captured by Asite/Ain. Both metrics are physically interpretable and rely solely on digital elevation data, making them broadly applicable to other cold regions. These findings provide a unified and mechanistically grounded framework for understanding extreme nocturnal cooling across diverse terrain types and offer practical tools for identifying areas vulnerable to extreme low temperatures in cold-region agriculture and forest management.
