JAUE2017-065: Investigation on Temperature Resolution Enhancement in Temperature Sensitivity -Improvement of Population Weighted Average Temperature using WRF operation and Temperature Sensitivity analysis in August

Abstract

The purpose of this study is to investigate whether Weather Research & Forecasting (WRF) data is useful in 
considering temperature sensitivity. The population weighted average temperature(PW-T) previously calculated, 
hereafter referred to as PW-T(AMeDAS), was based on measured values of AMeDAS. Recently with the help of 
WRF, we increased the number of data by interpolation calculation and improved the resolution of population 
weighted average temperature, hereafter referred to as PW-T(WRF+AMeDAS). For the period of August 2010, 
PW-T(WRF+AMeDAS) was 0.63℃ higher at maximum than PW-T(AMeDAS). This small difference does not 
deny the results we obtained before with PW-T(AMeDAS), but PW-T(WRF+AMeDAS) represents wide area 
average temperature better. After regression analysis of temperature sensitivity for August 2010, we found the 
coefficient of determination calculated by PW-T(WRF+AMeDAS) is greater than that calculated by 
PW-T(AMeDAS) for those days of holidays. For weekdays, the coefficients in both cases are high enough, 
showing a strong correlation between power supply and temperature. At a certain temperature, the power supply 
calculated by PW-T(WRF+AMeDAS) is lower than that calculated by PW-T(AMeDAS). It is suggested that 
PW-T(WRF+AMeDAS) could be an effective parameter to be used for temperature sensitivity analysis.