基于DMSP/OLS夜间灯光指数的浙江GDP空间化

Spatialization of GDP in Zhejiang Province Using DMSP/OLS Night Light Data

为了突破GDP数据在研究中受到的行政区单元限制,获得更加精细的GDP分布情况从而实现对社会经济的状况的深入探索,以DMSP/OLS夜间灯光影像数据为基础,进行了浙江省GDP空间化。采用地理与数理统计相结合的办法,获取2013年浙江省夜间灯光各项指标数据与GDP数据进行各类回归分析,选取最佳拟合方式作为GDP预测模型(R^2 = 0.965),检验结果发现各地级市预测GDP与实际统计值相对误差小于0.1,具有可用性。并根据上述结果获取GDP密度图,将GDP水平分配到每个11KM像元,实现GDP空间化直观表达。结合表达结果,打破行政统计单元局限,细致分析了浙江省经济分异情况,并给出了相关建议。

In order to break the limitation by administrative unit through GDP study,get more elaborate distribution of GDP,so as to realize the further exploration of social and economic conditions,use DMSP/OLS night lighting image data as the foundation,carried out the spatialization of GDP of Zhejiang province.By adopting the combination of geographic and mathematical statistics method,get night lighting indicators data in Zhejiang province,2013,then,carried on several kinds of regression analysis with GDP data,to select the best fitting way as GDP forecast model(R^2 = 0.965),the inspection found that the error is less than 0.1,with availability.Retrieve GDP density figure according to the above results,the level of GDP allocated to each 1x1 km pixel,achieve spatialization of GDP visual expression.Combining GDP density figure,break the administrative statistics unit,detailed analysed the economic differentiation in Zhejiang province,at last,this paper gives some related suggestions.