Local Climates of Yamaguchi Valley (I) : Statistical Analysis for Forecasting Night Minimum Temperature

The authors tried to forecast night screen minimum temperature at Yamaguchi Observatory by the statistical methods using some meteorological data at 1500hr of 197 days from 1966.4.1. to 11.30, eliminating the days on which the warm or cold front passed over Yamaguchi Valley. The result are shown on Fig. 1 and 2. Fig. 1 shows the relation between ΔT(℃) (=air temperature at 1500hr, T_<1500> － night minimum temperature, Tmin) and relative humidity, f(%)at 1500hr. Solid line is drawed according to the regression equation, ΔT=e^<0.0338(100-f)+0.7578>, which was fitted by the method of least squares, and dotted lines show the confidence limit with the probability 95% of the calculated value <ΔT>^^^, Fig. 2 shows the relation between observed ΔT and calculated one by a multiple regression equation, ΔT=31.44-0.211f_<1500>－0.415T_<1500>, and correlation coefficient between them is 0.57 less than the critical value 0.7 worth forecasting. A physical method which considers the flow of cold air in this valley, outgoing radiation at night, and ground surface conditions will be discussed in the near future.