It is one of the essential climatological investigations to classify temperature fields spatially continued at the surface, i. e., to set up climatic boundaries in the fields. In many investigations isothermal maps drawn from sample means have been practically used to set up boundaries in the fields. There is a possibility that the isothermal maps are different from one sample to another which are even taken under the same conditions, respectively. Therefore, classification of temperature fields might be different from one sample to another. In this investigation the author classified winter temperature fields in Kanto Plain, central Japan, using inferential statistical methods. The study area is divided into units of regular triangles. Each apex of the triangles corresponds to a point at which temperature is analysed. Sample means of temperature at three points of each unit are tested one another by F and “student-t” tests to examine whether population mean values of temperature at the points are significantly different one another. Then two areas are set up. One is continuous area which is a unit of population mean values at the points being not significantly different one another. The other is discontinuous area which is a unit of more than a pair of the population mean values being significantly different one another. Temperatures in discontinuous area, therefore, are different for different places. From the results of the tests, each unit is identified with continuous or discontinuous area. Triangle sides between continuous and discontinuous areas are here regarded as climatic boundaries. Above-mentioned method of classifying surface temperature fields was applied to winter temperatures (daily maximum, minimum and range of temperature) in the plain. Data of temperature were taken for the winters (December, January and February) from 1961 to 1964 at seventy-eight observation stations distributed at almost equal distances in the plain. The study area was divided into units as shown in Fig. 2. Based on the comparative examination in synoptic climatological investigations (Hohgetsu, 1978, 1979), the criterion of the surface pressure patterns in the Far East (Yoshino and Kai, 1974) was adopted here as the conditions under which samples were taken. The author analysed surface temperature fields of winter monsoon, troughs and migratory high patterns of the criterion, referring to relative occurrence frequencies of the patterns (Table 1). As far as maximum temperature is concerned, boundaries between continuous and discontinuous areas appear in the marginal parts of the plain, and continuous areas occupy broadly the middle parts of the plain. Regarding minimum and range of temperature, distribution of continuous areas under each pressure pattern are shown in Figs. 3 and 4.
Read full abstract
Dec 31, 1986
O.B Olesen 
Open Access
