云南沿边地区1992-2013年GDP空间化数据集

Gridded GDP Dataset of Yunnan Border Area(1992-2013)

云南沿边地区地处云南省的西南边陲地带,处于缅、印、中、孟经济走廊,具有重要的地缘位置。因此,较高精度地实现云南沿边地区国内生产总值(Gross Domestic Product,GDP)空间化拟合显得尤为重要。本研究以云南沿边地区为研究区,以DMSP/OLS夜间灯光数据、土地利用数据、云南省统计数据等为数据源,实现1992-2013年研究区的GDP空间化拟合。对夜间灯光数据进行饱和校正、相互校正、年内融合、年际间校正、重投影、重采样及裁剪等处理,基于土地利用数据实现第一产业的空间化,基于夜间灯光数据,采用"分类回归"的方法实现第二、三产业的空间化,从而得到GDP的空间化,并且对拟合结果进行验证。结果表明:各期的第一产业拟合值相对误差最大为1.12%,各期误差均较小;第二、三产业经过分类回归后的拟合相对误差均不高于-6.40%,最终GDP拟合的相对误差每期误差均小于-4.24%,精度较高。该数据集存储为.tif格式,单个文件的空间分辨率为1 km,共22组文件,数据量为68.6 M(压缩为1个文件,3.43 MB)。

The Yunnan border area is located on the southwestern border of Yunnan Province. It is located in the economic corridors of Myanmar, India, China, and Bangladesh and has an important geographical position. Therefore, it is critical to implement the spatial fitting of the Gross Domestic Product(GDP) data in the Yunnan border area with high precision. In this study, the Yunnan border area was used as the research area, and DMSP/OLS nighttime light data, land use data, and Yunnan provincial statistical data were used as data sources to implement the spatial fitting of the GDP data for the research area from 1992 to 2013. Saturation correction, mutual correction, annual fusion, interannual correction, reprojection, resampling, and clipping were performed on nighttime light data. Spatialization of the primary industry based on land use data was implemented. Based on nighttime light data, a "classification regression" method was used to implement the spatialization of the secondary and tertiary industries to obtain the spatialization of the GDP and verify the fitting results. The results show that the error of the primary industry fitting value of all periods is 1.12% at the maximum, and the fitting error is small. The relative error of the fitting of the secondary and tertiary industries after classification regression is less than 6.40% in each period, and the relative fitting error of the final GDP of each period is less than 4.24%, with high accuracy. The dataset is stored in the.tif file format. The spatial resolution of a single file is 1-km, and there were a total of 22 groups of files;the amount of the data volume was 68.6 MB(when compressed to 1 file, it was 3.43 MB).