Abstract
Most of the AMeDAS (Automated Meteorological Data Acquisition System) stations measure temperature, precipitation amount, sunshine duration, wind direction and velocity. However, no measurement on vapor pressure is made, although it is one of the indispensable meteorological elements for the estimation of air-conditioning load. This paper reports the investigation of the possible methods for appending hourly vapor pressure to AMeDAS data. Mainly two methods have been proposed and examined. A brief explanation on these two methods and the accuracy of the appended vapor pressure are as follows; 1) Method-1: This method predicts hourly vapor pressure at an arbitrary AMeDAS station by modifying the temperature, measured at the same station. When the annual and daily periodic components are extracted from the time series of hourly temperature for a year, the variation of the residual component becomes similar to that of the time series of hourly vapor pressure of the same year, provided that the annual periodic component is extracted. Therefore, the time series of hourly vapor pressure for a year can be predicted, by combining the residual component of the temperature and estimated annual periodic component of the vapor pressure. The estimation of errors by this method is around 2.5hPa. 2) Method-2: This method appends the hourly vapor pressure to an arbitrary AMeDAS station by interpolating the hourly vapor pressure observed at the surrounding weather stations. Several interpolation methods are examined and it clarifies that the weighted average of the observations by the reciprocal of distances between AMeDAS and the weather stations, gives a reliable prediction for the vapor pressure at the AMeDAS station. However, adjustments based on the altitude and temperature difference between the AMeDAS and weather stations are required when the temperature and the altitude differences exceed to a certain degree. The estimation of errors by this method is around 1.2hPa.
