利用夜光数据的东北三省国内生产总值预测

Prediction of GDP in Northeast China Based on Nighttime Light Data

针对GDP的传统统计方法较繁琐、工作量大且存在滞后性的问题,提出了一种基于夜光遥感影像的GDP获取方法。首先通过掩膜提取DMSP-OLS数据得到稳定灯光影像集,其次对数据集进行平稳化处理、去饱和处理等,建立时间序列数据集,最后分析时间序列数据集与GDP之间的对应关系,建立多项式GDP预测模型。研究表明,该方法操作简单、预测时间短、效率高,具有较高的精度和预测吻合度,能够较准确反映研究区的GDP实际增长情况。

Aiming at the problem that the traditional statistical method of GDP is cumbersome,heavy workload and lagging,a GDP acquisition method based on nighttime light remote sensing image is proposed.Firstly,the DMSP-OLS data is extracted through the mask to obtain the stable lights image set.Secondly,the data set is smoothed and desaturated,and the time series data set is established.Finally,analyze the correspondence between the time series data set and GDP,and establish the polynomial GDP forecasting model.The research shows that this method has advantages of simplicity,short time and high efficiency,and has high accuracy and predictive coincidence,which can accurately reflect the actual GDP growth in the study area.