Abstract
In order to estimate the daily snowfall amount on the Japan Sea side of the Japanese islands in the winter season, statistical analysis and statistical method for estimation of the daily snowfall amount at each station in Niigata prefecture are discussed. The various distribution functions fitting to the observed frequency distribution at each station are discussed. They follow Poisson distribution function on the coast of the Japan Sea, polya-type distribution in the mountain districts and exponent-type distribution in the intermediate districts. As they are too far from the normal distribution function, it is suggested that some treatment, such as the discriminant analysis method, is desirable. The statistical relationships of the classified daily snowfall amount and various meteorological data at Wazima are discussed. The statistical study of severe snowfall indicates extremely low temperatures and instability from the ground surface to the middle layers of the troposphere and higher temperatures and stable layers in the upper levels. In the case of moderate snowfall the condition reverse. In the lower parts of the cold dome, the vertical transfer of sensible heat and water vapor from the sea surface is remarkable. It was also shown that such a situation exerts an influence on the Bulk-Richardson-Number in PBL. Discriminant functions for the classified daily snowfall amount at each station are contsructed from the above statistical analysis. Using independent data, verification of the acuracy of the discriminant functions is also made. This discriminant method is found to afford more information than the conventional climatological method.
