| With the acceleration of industrialization and urbanization process in China,carbon emissions have increased significantly.In order to implement emission reduction measures effectively,it is necessary to developing a method for consistent and normative dataset of carbon emissions at fine spatial scales.Combined with the nighttime light data and gridded population data,this paper used an improved distribution model of carbon emission to map the spatial distribution of carbon emissions in Guangdong Province in 2005,2010 and 2013,and analyzed the spatiotemporal variations of carbon emissions.The main conclusions of the study are as follows:(1)Based on the region with stable DN value,the effective nighttime light data could be obtained by intercalibration of nighttime light images acquired by different sensors,and the saturation state of nighttime light data was removed based on the negative correlation between MODIS EVI and nighttime light data.Using the region with stable DN value to construct a quadratic linear equation,we corrected the nighttime light images acquired by different sensors and obtained comparable nighttime light data with continuous time series;According to the linear relationship that with the urban area vegetation index increased,the nighttime light brightness value decreased,the adjustment coefficient was constructed to perform nighttime light data desaturation correction.After the correction,the saturated pixel DN value is close to the actual light value,and the change characteristic of the light value in city center area is highlighted.(2)An improved distribution model of carbon emissions was performed and the spatial distribution of carbon emissions at the grid scale was obtained.The improved carbon emission distribution model,which was applicable at the city level,used the per capita carbon emissions ratio parameter representing the difference between rural and urban carbon emissions.This study mapped the spatial distribution of carbon emissions in Guangdong Province in 2005,2010 and 2013 at grid scale,by combining nighttime light imagery and gridded population data.The results showed that the carbon emissions were effectively distributed to grid cells.The carbon emissions with unlighted values were distributed based on the gridded population data,which solved the limitation of traditional regression method to estimate the carbon emissions of unlit area as zero.(3)This paper concludes that the spatiotemporal variations of carbon emissions are obvious at both pixel level and county level in Guangdong Province.Through analyzing the spatial and temporal distribution characteristics of carbon emissions in Guangdong Province,it can be seen that the carbon emissions at the county level are generally on the rise,and the largest carbon emissions at the pixel scale are in the Pearl River Delta region.Global Moran’s I and Hotspot analysis were used to test the spatial autocorrelation of carbon emissions in Guangdong Province,and it was found that carbon emissions in the Pearl River Delta region had a high high clustering feature,while carbon emissions in eastern Guangdong Province had a low-low clustering feature.The spatiotemporal heterogeneity of carbon emissions indicates that economic development of Guangdong Province was extremely unbalanced.The research indicates that spillage effects should be taken into account when formulating carbon emission reduction policies and measures. |
