夜间灯光遥感数据的GDP空间化处理方法

GDP Spatialization in China Based on Nighttime Imagery

随着夜间灯光遥感数据的应用日渐成熟和资源环境研究领域,对空间型社会经济数据的需求增加,利用相关分析和回归分析的方法,首次定量探讨夜间灯光数据与统计型的社会经济数据的空间关系。为提高模型精度,按照我国省级行政边界分区建模,分析全国县级的地区生产总值、第一产业、第二产业、第三产业分别与夜间灯光指数的空间相关关系,最终建立全国的1km GDP密度图。结果表明,全国范围的夜间灯光数据与第一产业的相关性不明显,相关系数0.554,模型拟合效果差,R2为0.306;夜间灯光数据与地区生产总值、第二产业、第三产业均有明显的对数线性关系,尤其是与第二产业和第三产业之和,相关系数为0.824,R2为0.679。利用分区模型估算,生成的GDP密度图能较完整地反映全国社会经济分布详况,以及宏观分布特征。

Resources and environmental sciences require greatly the spatialized socio-economic data sets which are always obtained from administrative regions at national or provincial level,and accurate estimates of the magnitude and spatial distribution of economic activity have many useful applications.Developing alternative methods for making estimates of gross domestic product（GDP） may prove to be useful when other measures are of suspect accuracy or unavailable.Based on the summary and analysis of existing economic activity spatialization approaches and nighttime imagery applications in economic activity,this research explores the potential for estimating the GDP using relationship between the spatial patterns of nighttime satellite imagery and GDP in China by correlation analysis and regression analysis using concerned data processing software.With the regional differences of China＇s economic development,logarithmic regression models have been established between different night light indexes and GDP,primary industry,secondary industry,tertiary industry and the sum of secondary industry and tertiary industry at the provincial level.A clear logarithmic linear relationship between nightlight imagery and GDP,especially the correlation coefficient of night light index and the sum of secondary industry and tertiary industry is 0.824 and R2 of them is 0.679 at national level,suggests that this method is available and feasible to estimate the spatial distribution of economic activity such as GDP.The result,1-km grid GDP map of China based on nighttime light data,by comparing with the other GDP spatialization approaches,shows the obvious advantage to reflect complete details and characteristics of the national secondary industry and tertiary industry distribution,which is extending the field of nighttime light data research and applications for the socio-economic data in resources and environmental sciences.