香港地区加权平均温度特征分析

Feature analysis of weighted mean temperature T_m in Hong Kong

利用香港King's Park探空站(站号45004)2003-2009年探空资料回归了大气加权平均温度Tm、地面温度Ts、气压es和水汽压Ps的线性公式.通过比较分析发现Tm-Ts单因素回归结果和Tm-Ts、es、Ps多因素回归结果没有显著差异,但基于本地化探空数据的回归公式精度比Bevis公式高;增加样本数回归分析并不能显著提高公式精度,采用最近一月探空数据回归公式即可很好地由Ts拟合下年Tm,拟合均方根误差(F-RMS)为1.946K;用2003年数据回归出的经验公式Tm=113.29+0.5863Ts去拟合2004-2009年的数据,拟合均方根误差基本没有差异,因此某地Tm-Ts经验公式一次回归可长期使用.通过对全国83个国际交换站2009年探空数据回归得出我国大陆地区最新Tm-Ts经验公式为Tm=53.244+0.783Ts,该一般公式拟合均方根误差与本站数据回归剩余均方根误差(RRMS)相当,可代替本地公式广泛使用.

We made linear regression to get the empirical formula of Tm and Ts,es,Ps using HongKong King’s Park sounding station（number 45004） data from 2003 to 2009,in which Tm means weighted atmospheric mean temperature,Ts means surface temperature,es means surface water vapor pressure and Ps means surface atmospheric pressure.It concludes that there is no significant difference between one-factor and multi-factor regression results,but the precision of regression formula based on local sounding data is higher than that of Bevis formula;the precision of regression formula can’t be significantly increased by annually adding sample numbers.It is sufficient to forecast Tm by Ts using sounding data of last month and the fitting root mean square is 2.031 K.Fitting root mean square（F-RMS） changes little when using Tm-Ts model（Tm=113.29＋0.586 3Ts） based on data of year 2003 to compute data from 2004 to 2009,so once the Tm-Tsmodel is set up,it can be used for long time.The Tm-Ts model（Tm=53.244＋0.783Ts） is set up using 83 exchanging sounding stations data of year 2009,which can be used to compute Tm of five stations from different climatic regions.It can be seen that the precision is almost the same with that computed using local data,so the Tm-Ts model（Tm=53.244＋0.783Ts） can be used widely in China.