Spatial and temporal variations in summer precipitation in Japanese mountain areas

Abstract

Abstract:We examined spatial and temporal variations in precipitation measured during summer season between 1976 and 2007 for 28 stations located in mountain areas across Japan using the amount of precipitation (Pr), the mean depth of precipitation events (Á), and the inverse of the mean interval times ( ). We obtained positive correlations between the period mean Pr (Pr) and the period mean Á (Á) and between Pr and the period mean ( ) for the 28 stations. Pr was more strongly related to Á than to , indicating the spatial variations in Pr that are primarily related to the variations in Á. In addition, Pr was more strongly related to Á than to for most stations on the basis of data for 1976-2007, indicating that the year-to-year variations in Pr are primarily related to Á. We also examined temporal trends in Pr, Á and for 1976-2007 and found no systematic trends for 23 of the 28 stations, suggesting long-term trends that are not common in mountain areas of Japan. The relationships between Pr and Á and between Pr and Á presented in this study enable us to generate a temporal precipitation distribution pattern based on only Pr and Pr data, respectively. Furthermore, probabilistic stochastic hydrological models require precipitation characteristics as input; thus, this study contributes to the determination of hydrological cycles and their possible future changes in Japanese mountain areas and therefore to water resource management.

Spatial and temporal variations in rainfall characteristics in mountainous and lowland areas in Taiwan

“…Note that the range of λ at mountainous stations (0.4–0.7, unitless) was smaller than that of lowland stations (0.2–0.7, unitless) (Figure a). Shinohara et al () also showed that η was the primary factor determining spatial variation in Pr over mountainous areas in Japan, but Dairaku et al () suggested that spatial variations in Pr in a mountainous region of northern Thailand were determined by rainfall frequency (i.e. λ ) rather than η .…”

“…In southwest China, Qin et al () examined trends of annual rainfall using long‐term rainfall data from 136 stations and found that 23 stations had an increasing trend and six stations had a decreasing trend. Shinohara et al () found no significant trends in rainfall characteristics for 23 of 28 stations in mountainous areas across Japan, and the stations with significant long‐term trends in these characteristics were scattered irregularly. These results suggest that long‐term trends in rainfall are highly localized, that is, each region has its own specific long‐term trends.…”

“…These parameters are useful not only for studying rainfall variability but also for hydrological applications such as stochastic models (e.g. D' Odorico et al, 2000;Porporato et al, 2004;Kumagai et al, 2009;Shinohara et al, 2010).…”

“…Rainfall characteristics in mountainous areas are often different from lowland areas (Dairaku et al , ; Shinohara et al , ; Qin et al , ). Takeshita and Mitsuno () and Shinohara et al () reported different long‐term trends in rainfall characteristics for rainfall stations within just 10 km of each other. These studies suggest that extrapolation of rainfall characteristics in mountainous areas from the results of Yeh and Chen () and Hsu and Chen () could be invalid and that spatial and temporal variations in rainfall and their long‐term trends are still unclear, especially in mountainous areas.…”

“…Areas with altitudes >1000 m and between 100 and 1000 m each account for about 33% of the area of the entire country (Teng et al, 2000;Lin et al, 2010). Rainfall characteristics in mountainous areas are often different from lowland areas (Dairaku et al, 2004;Shinohara et al, 2010;Qin et al, 2010). Takeshita and Mitsuno (2005) and Shinohara et al (2010) reported different long-term trends in rainfall characteristics for rainfall stations within just 10 km of each other.…”