LOCAL CLIMATOLOGY OF SNOW: DEPTH OF ACCU MULATION AND DIRECTION OF SNOWY WIND IN A COASTAL REGION BORDERING TOKYO BAY

Abstract

The effects of the distribution of land and sea on snowcover and direction of snowy wind in a coastal region is of striking interest in the local clizma tology of snow. The author attempted a survey in an area of about 20 square kilometers bordering Tokyo Pay. (Fig. 1) (1) As snowcover and terrain very from plane to place, it was necessary to determine an accurate method of snow depth measurement. For this purpose, special concideration was taken, as in Fig. 3; results are shown in Table 1. When the depth of snowcover is less than 30cm, the mean value for 5 measurements at a place have a variance of ±1 cm at the maximum and test results show that the values of 5 paints on a starlike line have a smaller variance than those on one straight line. (2) To determine the average depth of snowcover in this area, observa-tions were made early in the morning of Feb. 22, 1953 covered with snow which had accum ulated since the morning of Feb. 21st (Fig. 2). Results of these observations are shown in Fig. 4. .Depth is represented as a linear function of the logarithm of the distance from the coastal line, that is: y=a logx+b_??_ (1) wherex: depth of snowcover in cm, Y: distance from the coastal line in hm, a, b: constant. In this case a=11.49, b=1.09. (3) Applying the data accumulated at the 10 weather stations in this area, the value of the constant a in equation (1) was calculated fur each of 20 cases. As is shown in Table 2, a is positive with but one exception. This means that the snowcover, in general, increa=yes in depth with the distance from the coastal line. It maybe explained by the effect of diffe-rences in air temperature and wind velocity between the seaeoa dt and inland points. (4) The constant a is not related to the depth of snowcover, but rather to the length of the duration of snowfall. The exceptional cases seem to be related to the influence of the front remaining at the eoast and bringing frontal snowfall for a relatively long period. The experimental equation is a=0.36 t-0.12_??_(2) where a :constant in equation (1), t: duration of snowfall in hours. In theoretical considerations, if t becomess 0, a must be 0, and practically the second term is smaller than the first _??_because t must be greater than 4 (hours), the shortest time studied in this paper_??_. Therefore, as the second term is negligible, it can be approximated as: a=0.36t_??_(21) Then, y=(0.36t) logx+b_??_(3) where y: depth of snowcover in cm, x: distance from the coastal line in hm, t: duration of snowfall in hours, b: constant determined by meteorological conditions in each case. (5) The prevailing wind direction during the snowfall was also observed at 136 points in the morning of Feb. 22, 1953 by utilizing the fact that when air temperature is near 0°C, falling snowflakes tend to adhere to the windward surfaces of electric poles and tall chimneys. Result s, as shown in Table 3, indicate the prevailing snowy wind direction to have been N y NAT at the inland stations gradually turning to I at a distance of about 8 k ml from the seaccoast where it is almost due N. This is considered to be a result of the difference in the time of snowfall occurrence as affected by the different air temperatures. in this area or of the difference of friction at land and sea.