基于夜光遥感的GDP空间化建模及动态分析——以河南省为例

GDP spatial model and dynamic analysis based on night-time light imagery:a case from Henan

针对GDP数据统计通常基于行政单元无法精细表达行政单元内部的GDP空间分布及其时空动态演变的情况,使用2006、2010和2013年DMSP/OLS灯光数据计算了河南省夜间灯光指数,分析不同灯光指数与GDP的线性相关性构建了GDP的最佳回归模型,并利用相对应年份的统计数据进行精度验证,实现了河南省GDP空间可视化表达及动态分析,制作了1 km×1 km的GDP密度分布图。结果表明:2006、2010和2013年的夜间灯光总强度指数TNL与GDP的空间分布线性相关性明显,R^2分别为0.9659、0.9766、0.9790;各行政区的GDP模拟值与统计值的平均相对误差分别为8.19%、6.04%、8.08%。2006—2013年河南省18个地级市GDP密度图呈现出明显的扩张态势,模型精度高,表明灯光数据可作为评价河南省经济状况的重要指标。

Data statistics for GDP are usually based on administrative units.However,it can not express the GDP spatial distribution and inside dynamic spatial and temporal variations of administration units accurately.By DMSP/OLS night-time light data in 2006,2010 and 2013,the night light index is calculated.The linear correlation between different light index and GDP is analyzed,and the best regression model of GDP is constructed.The accuracy by the statistical data of 2006,2010 and 2013 is verified,the spatial visualization expression and dynamic analysis of GDP in Henan are realized.A map of 1 km×1 km GDP density distribution is made.The result shows that there is a significant linear correlation between total night-time light index and the spatial distribution of GDP in 2006,2010 and 2013.The values of R^2 are 0.9659,0.9766 and 0.9790,respectively.From 2006 to 2013,the GDP density maps of 18 prefecture-level cities in Henan showed an obvious expansion trend.The model of this study with high accuracy proves that the light data can be used as an important index for estimating and analyzing the economic situation of Henan.The model precision of this study is comparatively satisfied.