Abstract
In this paper, the statistical formulae for predicting the vertical gradient of potential temperature, available for the layer about 15 to 125 meters in height, were developed. Parameters for these formulae consist of hourly surface level wind speeds and screen-height temperatures, which can be easily obtained from routine meteorological data.
The estimated vertical gradients were checked their efficiency by the following methods: one was the comparison with the observation at the meteorological tower, and the other was with the concentration of carbon monoxide. In the former case, the proposed method shows better agreement with the observation than the statistical estimates derived from the Pasquill-Gifford stability class. In the latter case, the estimated vertical gradients were averaged all over the stations in Tokyo. And after this, these data were compared with the area-averaged concentration data. The result also shows the improvement in the correlation coefficient, compared both with the case using the observed temperature gradients at each tower and with the one using the Pasquill-Gifford stability class.
