STEVE气温预测模型的普适性检验——以广州地区为例

Validation Tests for Air Temperature Prediction Model STEVE:An Example of Guangzhou

STEVE(Screening Tool for Estate Environment Evaluation)是一种经过回归分析得出的统计模型,通过获得气候变量和城市形态变量,能够对新加坡当地的气温指标进行快速预测,但该模型的普适性并没有得到检验,不能直接应用于新加坡以外的地区。通过3组检验过程,即对介绍STEVE的原文献中的数据进行易地验证,以及在广州地区进行2组实测实验验证,证明STEVE能够较好地满足工程上的精度要求,可以作为一种简易预测气温的工具模型应用于广州地区以及具有相似气候背景城市的规划与设计阶段。STEVE还具有优化完善的空间,在未来的研究工作中将继续扩展其适用范围。

STEVE（Screening Tool for Estate Environment Evaluation）is a statistical model through regression analysis.It can predict the air temperature by climate and urban morphology predictors for Singapore estate,while it hasn＇t been proved valid in other areas.Consequently,three groups of validation tests in Guangzhou area were conducted,one using the original data in the literature and the other two using experimental data from the field measurement.It is shown that STEVE tool can meet the engineering accuracy and can be used as a simple temperature prediction model tool in urban planning and design process in Guangzhou area and other cities with similar climate conditions.Besides,the STEVE tool can be optimized further to expand its applicable scope.