基于多元线性回归模型的华南地区≥10℃积温空间模拟研究

摘要

选用1980～2011年华南地区111个及周边地区39个气象站地面观测资料，基于ArcGIS10.0平台，利用多元线性回归模型和IDW、 Kriging、 Spline等插值方法模拟了华南地区≥10℃活动积温的空间分布。结果表明：基于多元线性回归模型的≥10℃活动积温空间模拟方法与IDW、 Kriging、 Spline等插值方法相比，解决了积温空间分布模拟复杂性的问题，并考虑了≥10℃活动积温受海拔、经纬度差异等因素的影响，具有较高的模拟精度，适合于较大范围的积温数据空间化模拟。另外，1980～2011年华南地区≥10℃活动积温随海拔的升高而减小；随纬度的增大而减小；随经度的变化幅度较大，在研究区西部云南地区受地形的影响变化幅度强烈，在研究区东部广西、广东、福建等地变化幅度较小。%The spatial distribution of active accumulated temperature ≥10℃ (AAT10) was simulated by using the multiple linear regression model (MLRM) and IDW, Kriging, Spline interpolation method based on the daily meteorological observation data from 111 meteorological stations in Southern China and surrounding areas during 1980 to 2011. The result shows that compared with the conventional methods such as IDW, Kriging and Spline, it solves the complexity of the spatial distribution of AAT10, and considers the elevation, longitude and latitude differences to AAT10. It shows higher simulation accuracy, and is suitable for a wide range of spatialization of accumulated temperature. At the same time, the AAT10 in South China was reduced with increasing altitude from 1980 to 2011; and reduced with increasing latitude. There is a large variation with longitude in the western Yunnan region affected by the impact of changes in topography, and a little variation with longitude in the eastern part of Guangxi,Guangdong, Fujian province.