基于卫星夜间灯光数据的中国分省碳排放时空模拟

中国能源统计数据"横向不可比,纵向不可加"现象依然突出,尤其是分省能源消费统计千差万别,给分省碳排放评估带来了较大困难,如何利用卫星遥感数据科学合理地估算中国分省碳排放是当前亟须研究的问题。本文运用DMSP/OLS全球稳定夜间灯光数据,在通过相互校正、年内融合和年际间校正等系列处理得到中国分省稳定夜间灯光数据的基础上,首先分别构建中国分省稳定夜间灯光亮度DN值与人均碳排放和单位面积碳排放之间的时空地理加权回归模型,两个模型整体效果均较好,拟合优度分别高达96.74%和99.24%;其次运用稳定夜间灯光亮度DN值对分省人均碳排放和单位面积碳排放进行时空模拟;最后运用人口规模和土地面积对分省碳排放进行估算。估算结果显示:(1)整体来看,2000—2013年年均碳排放模拟值与实际值6.3349×109t较为接近,两个模型的相对误差均在0.5%以内。(2)分年度来看,所有年份的相对误差均在5%以内,2006年分省加总碳排放模拟值与实际碳排放6.2036×109t最为接近,绝对误差和相对误差均较小,两个模型模拟值的相对误差均为0.04%。(3)分省域来看,2000—2013年年均碳排放模拟值与实际碳排放均非常接近,除海南和宁夏外,其余28个省区市的相对误差均在1%以内。(4)分年度分省域来看,以2013年为例,40%省份的相对误差在2%以内,70%省份的相对误差在5%以内。从整体、分年度、分省域、分年度分省域的估算结果来看,基于稳定夜间灯光数据的中国分省碳排放时空模拟效果良好。因此,运用卫星夜间灯光数据可以较为准确地对中国分省碳排放进行估算和预测,为卫星遥感影像数据服务分省碳排放监测和评估提供一种补充性参考。

Spatiotemporal simulation study of China's provincial carbon emissions based on satellite night lighting data

China's energy statistics'horizontal is not comparable,vertical cannot be added'problem is still outstanding,especially the provincial energy consumption statistics vary widely,brings great difficulties to the provincial carbon emissions assessment.How to use satellite remote sensing data to estimate China's provincial carbon emissions scientifically and reasonably is an urgent problem to study.This paper,based on the stable night lighting data of different provinces in China obtained by using series correction such as intercalibration,intra-annual composition,inter-annual series correction to DMSP/OLS global steady night lighting data,first,constructs the geographically and temporally weighted regression model between the night DN value and the per capita carbon emission,and the carbon emission per unit area.The two models have good overall effect and the goodness of fit is 96.74％ and 99.24％ respectively.Second,it uses the DN value of steady night lighting intensity to simulate the carbon emission per capita and the carbon emission per unit area.Finally,it uses the population size and land area to estimate provincial carbon emissions.The results show that:①On the whole,the annual carbon emission simulation value from 2000 to 2013 is close to the actual value of 6 334.9 million tons,the relative error of the two models are within 0.5％.②On a year-by-year basis,the relative error of all years is within 5％,the total simulated carbon emissions value of provinces in 2006 is closest to the actual carbon emissions of 6 203.6 million tones,the absolute error and relative error are small,and the relative errors of the two models are beth 0.04％.③In the provinces,the average annual carbon emissions in 2000-2013 is very close to the actual carbon emissions,with the exception of Hainan and Ningxia,and the remaining 28 provinces relative errors are less than 1％.④In the case of 2013,for example,the relative errors of 40％ of the provinces are less than 2％,and the relative errors of 70％ of the provinces are less than 5％.Based on the whole,the annual,the sub-provincial,and the sub-provincial sub-annual estimated results,the simulation results of carbon emissions in China based on the data of stable night lighting are good.Therefore,using satellite night lighting data can estimate and forecast China's provincial carbon emissions accurately,and provide a supplementary way for using the satellite remote sensing image data to monitor and assessing the sub-provincial carbon emissions.