基于DEM模型和气候学计算的滇南山区太阳总辐射空间化

Total Solar Radiation Spatialization in Mountainous Region based on DEM and Climatological Calculation

地形和地势引起的温度、热量、水分差异是局地自然地理环境空间格局分异的内因所在，而太阳总辐射是地表物理、化学、生物等过程的主要能源，是理解或解释陆地表层格局的重要基本参数。本研究结合太阳总辐射的两种主要气候学计算模式Angstrom模型和Bristow—Campbell模型，以DEM、月均温差、月平均日照百分率以及地理纬度、赤纬等为基础数据，通过潜在太阳总辐射的空间化实现和太阳总辐射的气候学推算两个过程，成功实现了云南省中南部山地的多年平均1月～12月太阳总辐射的空间化；通过两个有多年太阳总辐射数据的站点实测数据检验，模拟结果良好：各月太阳总辐射平均误差率均小于10％，平均为3．69％；同时，也提供了一个可以借鉴到其它山地丘陵区的太阳总辐射空间化模式。

Different temperature, precipitation and solar radiation, resulted from different regional landform and elevation, are the main factors of regional differences. Total solar radiation is the major energy source of earth surface processes, including physical, chemical and biological processes, and also a basic parameter to understand or explain the changes of land surface patterns. In this research, we build a multi-year average model for annual total solar radiation spatialization from 1961 to 2000 by using two main climatological calculation models, Angstrom model and Bristow-Campbell model, at the same time. It has been realized through two processes, including both the spatialization of potential solar radiation and the climatological calculations of solar radiation between different models, and has successfully resolved the spatialization of monthly total solar radiation databases with an area of 250m x250m in southern Yunnan Province, regarding DEM, monthly differences of minimum and maximum temperatures, latitude, declination, and monthly average sunshine percentage as the basic data. More importantly, we find the simulation result quite good by contrasting land-ground measurement data and simulated grid data. Solar radiation error rates of each month are 3.69% on average, far below 10%. Meanwhile, this applicable spatialization mode of total solar radiation in southwest of China can be used in other regions for reference, especially in mountainous areas. In the application of this model, two aspects can influence error ratio. The first one is how to decide the parameter value of surface vegetation reflectance in different months because there is an obvious seasonal change of vegetation cover throughout a year. It may be the main source of error in this model and is one of the key problems to be resolved in the next step of our research. The second one is how to improve the DEM spatial resolution. The error ratios may be much lower if the spatial resolution of 250m × 250m in this study can be improve to 100m × 100m or 25m × 25m since the largest surface characteristics of the mountainous region is the complicated terrain.