资源科学 ›› 2019, Vol. 41 ›› Issue (6): 1131-1140.doi: 10.18402/resci.2019.06.12

• 气候资源 • 上一篇    下一篇

考虑太阳辐射修正的重庆山地气温空间化模拟

何志明1,2(), 李月臣3,4(), 金贤锋1,2, 刘贤1,2, 何小波1,2   

  1. 1. 重庆市地理信息中心,重庆 401147
    2. 时空大数据技术研究与应用重庆市工程实验室,重庆 401147
    3. 重庆师范大学地理与旅游学院,重庆 401331
    4. GIS应用研究重庆市高校重点实验室,重庆 401331
  • 收稿日期:2018-11-21 修回日期:2018-12-24 出版日期:2019-06-25 发布日期:2019-06-25
  • 作者简介:

    作者简介:何志明,男,山东德州人,硕士,工程师,主要研究方向为遥感与GIS应用研究。E-mail:hzm@dl023.net

  • 基金资助:
    重庆市技术创新与应用示范专项社会民生类重点研发项目(cstc2018jscx-mszdX0067);国家自然科学基金项目(41571419);国家重点研发计划(2018YFB0505400)

Spatial interpolation of mean temperature of Chongqing Municipality considering solar radiation correction

Zhiming HE1,2(), Yuechen LI3,4(), Xianfeng JIN1,2, Xian LIU1,2, Xiaobo HE1,2   

  1. 1. Chongqing Geomatics Center, Chongqing 401147, China
    2. Chongqing Engineering Laboratory of Spatio-temporal Big Data Technology Research and Application, Chongqing 401147, China
    3. School of Geography and Tourism, Chongqing Normal University, Chongqing 401331, China
    4. Key Laboratory of GIS Application, Chongqing Municipal Education Commission, Chongqing 401331, China
  • Received:2018-11-21 Revised:2018-12-24 Online:2019-06-25 Published:2019-06-25

摘要:

为实现重庆多山、多云雾、少日照等典型地理环境特征下气温的空间分布精细化模拟,本文提出了一套局部回归加地形影响修正的适宜性模型方法。该方法综合地理加权回归模型、Solar Analyst模型、改进的Angtrom-Prescott方程以及多元线性回归,基于气象站观测的气温、相对湿度、日照百分率参数以及辐射站太阳总辐射参数,结合100 m×100 m DEM数据,进行山地起伏地形下气温空间化模拟。其中,气温的地形影响修正通过起伏地形下太阳总辐射的拟合而实现。模型具有较好的模拟精度和稳定性,局部回归项的模拟精度远高于反距离权重插值(IDW)、克里金插值(Kriging),也总体优于传统的基于纬度、经度、海拔高度、日照百分率、相对湿度因子构建的全局多元回归模型;采用55个区域气象站进行单一年份夏季气温模拟精度验证,平均绝对误差为0.59℃,地形影响修正后有38个站误差降低。模型具有较好的时空维度模拟能力,能反映坡度、坡向、地形遮蔽等局地地形因子对气温的影响,具有较强的物理意义。模型与商业化的ArcGIS软件工具相结合,便于推广应用,特别适用于重庆及其周边西南山地太阳辐射低值区。

关键词: 气温插值, 太阳辐射修正, DEM, 山地气温, 重庆

Abstract:

The mountainous regions of Southwest China, where Chongqing Municipality is located, has typical regional environmental characteristics such as cloudy fog and less sunshine. In order to realize the spatial simulation of temperature in this geographical environment, this study proposes a model for local regression considering terrain correction factor for solar radiation. In this model, the terrain correction factor is derived indirectly by fitting the spatial distribution of global solar radiation under undulating terrain. The model combines the Geographically Weighted Regression model, the Solar Analyst model, the improved Angtrom-Prescott equation, and the multiple linear regression method. Based on temperature, relative humidity, sunshine percentage, and global solar radiation of the meteorological stations, combined with DEM data with a resolution of 100 m×100 m, this model realizes the spatial simulation of temperature under the mountainous terrain. The model has good fitting accuracy and stability. The simulation accuracy of local regression term is much higher than Inverse Distance Weighting (IDW) interpolation and Kriging interpolation. It is also better than the traditional Multivariate Llinear Regression model based on latitude, longitude, altitude, sunshine percentage, and relative humidity. Further, 55 regional meteorological stations are used to verify the summer temperature simulation accuracy of a single year. The average absolute error is 0.59°C, and the errors of 38 meteorological stations are reduced after considering the terrain correction factor. The model performs well in spatial and temporal simulation of air temperature, which can reflect the influence of local terrain factors such as slope, aspect, and topographic occlusion on temperature, and has clear physical meaning. Based on the available observation data of meteorological stations, DEM, and the commercial software ArcGIS, this model is convenient to apply, especially suitable for cloudy, sunless areas like Chongqing and its surrounding mountainous regions.

Key words: spatial interpolation of temperature, solar radiation, DEM, mountain temperature, Chongqing Municipality