1/√N則によるアンサンブル平均日射変動の標準偏差算定に適した地理的範囲

抄録

In this study, we evaluated the practical usability of so-called 1/√N rule, focusing on three factors, i.e. the geographical extent, the location of representative point, and the topological difference. First, by using the insolation data observed at 37 points mainly located in the Nobi Plain, we calculated the correlation diagram between two different standard deviations (STDs) of geographical average insolation fluctuation, i.e. STD of ensemble average insolation of N points (σ_e) and STD calculated with the representative point data by using 1/√N rule (σ_N). As the results of evaluating the slope and R² values of regression line of correlation diagram and the P-value of χ²-test regarding frequency distribution of σ_e and σ_N, the 1/√N rule would be applicable for the area of around 20km radius. The representative point for calculating σ_N should be carefully selected even within the area of 15km radius. Because the difference in weather condition is much larger in the Matsumoto Basin area than in the Nobi Plain area even if the area size is the same, the 1/√N rule should be applied to smaller area than the area of 20km radius available for the Nobi Plain area.