基于SNPP-VIIRS夜间灯光数据的GDP预测模型对比  

Comparison of GDP prediction models based on SNPP-VIIRS nighttime light data

作　　者：詹淇雯 胡为安 刘传立[1] ZHAN Qi-wen;HU Wei-an;LIU Chuan-li(School of Civil Surveying and Mapping Engineering,Jiangxi University of Science and Technology,Ganzhou 341000,China;Guangdong Polytechnic of Industry and Commerce,Guangzhou 510510,China)

机构地区：[1]江西理工大学土木与测绘工程学院,江西赣州341000 [2]广东工贸职业技术学院,广州510510

出　　处：《桂林理工大学学报》2021年第4期843-850,共8页Journal of Guilin University of Technology

基　　金：国家自然科学基金地区基金项目(41561091);江西省教育厅科学技术研究项目(GJJ150663)。

摘　　要：由于基于夜光遥感数据的GDP预测多采用单一函数模型,缺乏对比和可行性验证,因此利用校正后的夜光影像对比多种函数模型的预测效果具有实际价值。鉴于此,选择SNPP-VIIRS作为基础数据,经过消除负值和背景噪声等校正处理,对线性函数等6种预测模型进行GDP估算效果对比。结果表明:在省级尺度下,二项式函数拟合优度最高,Gaussian函数预测效果最好;市级尺度下,二项式函数均优于其他函数模型。校正方法合理科学,校正后的影像可用于GDP预测研究。As a new type of data source,nighttime light data is used to record the brightness of night light on a global scale.It shows great potential and application prospects in estimating social and economic parameters.However,the GDP prediction based on remote nighttime sensing data mostly uses a single function model with out comparison and feasibility verification.It is necessary to use the corrected nighttime night image to compare the prediction effects of multiple function models.This study selects SNPP-VIIRS as the basic data.By correction processing,such as removing negative values and background noise,the GDP estimation effects of the six prediction models are compared,such as the linear function.Research indicates that the correction method is applicable and scientific.The corrected image can be used for GDP prediction research.At the provincial level,the binomial function has the bestfitting,and the Gaussian function has the best prediction effect.At the municipal level,the binomial function is superior to other function models.

关 键 词：夜间灯光数据 GDP SNPP-VIIRS 预测模型 拟合优度 

分 类 号：P407[天文地球—大气科学及气象学]