精准的人口空间分布信息，对于国家宏观政策的制定和国家战略的实施具有至关重要的作用。栅格尺度人口空间分布地图是研究生态环境变化、城市规划、区域协调发展、空气污染与流行病和灾害损失评估等的前提和基础。然而，传统的人口统计方法费时费力，精度低、更新缓慢且现势性较差，无法精确描述人口的实时动态分布特征，因此难以刻画行政区划内部的人口分布空间差异。夜间灯光遥感的出现为大尺度人口空间分布模拟提供了可能。然而，现有的基于夜间灯光数据的人口空间分布模型在不同时空尺度上的模拟精度存在不确定性，缺少基于多源夜间灯光数据，利用多种模型模拟人口分布精度的对比研究；同时，缺少栅格尺度上人口分布的时空动态研究。因此，本文选用索米国家极地合作卫星(Suomi National Polar Partnership，S-NPP)上搭载的可见红外辐射成像计(Visible Infrared Imaging Radiometer Suite，VIIRS)和珞珈一号（Luojia1-01）夜间灯光影像，利用多元线性回归（Multiple Linear Regression，MLR）、支持向量机（Support Vector Machine，SVM）、反向传播神经网络（Back Propagation Neural Network，BPNN）和随机森林（Random Forest，RF）开展了不同时空尺度的人口空间分布模拟与制图、精度对比和时空动态研究。

本文利用成渝城市群2013-2018年市、县和乡镇尺度的人口统计数据、不同尺度的行政区划边界数据、地形数据、土地利用数据、NPP-VIIRS夜间灯光影像、Luojia1-01夜间灯光影像、气象数据和社会感知数据，引入MLR和机器学习方法，包括SVM、 BPNN和RF，构建不同尺度上的人口空间分布估算模型，基于十折交叉验证并采用相关系数平方（R Squared，R²）、均方根误差（Root Mean Squared Error，RMSE）、平均绝对误差（Mean Absolute Error，MAE）和平均预测误差（Mean Predictive Error，MPE）指标评价模型精度以确定最佳模型；采用最佳模型对2013-2018年成渝城市群人口空间分布进行制图，得到了2013-2018年500×500m分辨率及2018年100×100m分辨率的人口空间分布地图产品，并采用已有人口空间分布栅格地图产品LandScan和Worldpop与本文所得产品进行交叉验证，评价人口空间分布地图的精度；利用所得产品开展了2013-2018年成渝城市群人口分布的时空动态特征研究。

主要研究结论如下：

（1）本研究基于NPP-VIIRS夜间灯光影像在成渝城市群市、县和乡镇尺度利用MLR、SVM、BPNN和RF模型在500×500m分辨率上进行了人口模拟。在不同方法的人口模拟中，RF的结果最好，市、县、乡镇尺度的R²分别为0.998，0.96和0.96，SVM的结果次之，市、县、乡镇尺度的R²分别为0.998，0.27和0.70，其次为BPNN，市、县、乡镇尺度的R²分别为0.996，0.23和0.66，MLR在市、县、乡镇尺度的R²分别为0.994，0.04和0.57；十折交叉验证中也是RF的结果最好，市、县、乡镇尺度的R²分别为0.996，0.87和0.88，其次为SVM，市、县、乡镇尺度的R²分别为0.95，0.20和0.66，MLR次之，市、县、乡镇尺度的R²分别为0.993，0.03和0.56，BPNN在市、县、乡镇尺度的R²分别为0.63，0.01和0.55。因此，选择RF作为最佳模型进行制图。

（2）利用RF在成渝城市群市、县和乡镇尺度对比了基于NPP-VIIRS夜间灯光影像2013-2018年和逐年人口模拟结果，选择了NPP-VIIRS利用RF模型进行市、县和乡镇尺度500×500m分辨率的人口空间分布制图，不同尺度的建模R²依次为0.998，0.96和0.96，验证R²依次为0.996，0.87和0.88；基于Luojia1-01利用RF模型进行市、县和乡镇尺度100×100m分辨率的人口空间分布制图，不同尺度的建模R²依次为0.97，0.81和0.93，验证R²依次为0.73，0.01和0.65。将2013-2018年500×500m人口空间分布地图与1km×1km的LandScan和Worldpop人口栅格化地图产品进行交叉验证，结果显示人口空间分布地图的准确性（R²=0.48）介于LandScan（R²=0.46）和Worldpop（R²=0.64）之间；利用2018年100×100m人口空间分布地图与Worldpop人口栅格化地图产品进行交叉验证，结果显示人口空间分布地图准确性（R²=0.38）低于Worldpop（R²=0.65）。

（3）利用人口空间分布地图，在栅格尺度对成渝城市群人口空间分布特征与时空变化规律进行了研究。结果显示：成渝城市群人口呈现西高东低，中间高四周低的分布特点，人口主要集中在西北部成都附近和中南部重庆附近，以人口密集区为中心向周围逐渐递减，两个人口密集区中间的四川盆地附近分布了中等密集区域；两个人口密集区的人口快速增长，人口较密区人口数量快速减少，四川盆地附近零星出现的人口密集区也出现较快增长趋势，其余大面积人口密度中等区、人口稀少区与人口极稀区的人口数量变化较小。

Accurate information on the spatial distribution of population is of vital importance to the formulation of national macro-policies and the implementation of national strategies. The grid-scale population spatial distribution map is the prerequisite and basis for studying ecological environment changes, urban planning, regional coordinated development, air pollution and epidemic and disaster loss assessment. However, traditional demographic methods are time-consuming and labor-intensive, with low accuracy, slow update and poor current status, and cannot accurately describe the real-time dynamic distribution characteristics of the population. Therefore, it is difficult to characterize the spatial differences in population distribution within administrative divisions. The emergence of nighttime light remote sensing had made it possible to simulate the spatial distribution of large-scale populations. However, the existing population spatial distribution models based on nighttime light data had uncertainties in the simulation accuracy of different time and space scales, and there was a lack of comparative studies based on multi-source night light data and the use of multiple models to simulate population distribution accuracy; at the same time, there was a lack of research on the temporal and spatial dynamics of population distribution on the grid scale. Therefore, this article selected the Visible Infrared Imaging Radiometer Suite mounted on the Suomi National Polar Partnership and Luojia1-01 nighttime light images, and used Multiple Linear Regression, Support Vector Machine, Back Propagation Neural Network and Random Forest to develop different time and space scale population spatial distribution simulation and mapping, accuracy comparison, and temporal and spatial dynamic research.

This article used the city, county and township-scale demographic data of the Chengdu-Chongqing urban agglomeration from 2013 to 2018, different-scale administrative division boundaries, topographic data, land use data, NPP-VIIRS nighttime light images, Luojia1-01 nighttime light images, meteorological data and social perception data, introducing MLR and machine learning methods to construct population spatial distribution estimation models on different scales, including SVM, BPNN and RF, based on ten-fold cross-validation and using correlation coefficient squared, Root Mean Squared Error, Mean Absolute Error and Mean Predictive Error as indicators to evaluate model accuracy in order to determine the best model; Then, this paper used the best model to study the spatial distribution of the population in the Chengdu-Chongqing urban agglomeration from 2013 to 2018. The products of the population spatial distribution map with a resolution of 500m in 2013-2018 and a resolution of 100m in 2018 were obtained, and the existing population spatial distribution grid products (LandScan and Worldpop) were used to compute the result of cross-validate with the products obtained in this article in order to evaluate the accuracy of the population spatial distribution map; Finally, the obtained products were used to carry out the temporal and spatial dynamic characteristics of the population distribution of Chengdu-Chongqing urban agglomeration from 2013 to 2018.

The main contents are as follows:

(1) Based on the city, county and township scale and NPP-VIIRS nighttime light images in the Chengdu-Chongqing urban agglomeration, this study used MLR, SVM, BPNN and RF models to conduct population simulations at a resolution of 500×500m. Among the population simulations of different methods, the RF results are the best, with R² at the city, county, and township scales being 0.998, 0.96, and 0.96, respectively. The results of SVM were second, with R² at the city, county, and township scales being 0.998, 0.27, and 0.70, respectively. Followed by the results of BPNN, R² at the city, county, and township scales are 0.996, 0.23, and 0.66, respectively. The MLR results were the worst, with R² at the city, county, and township scales being 0.994, 0.04, and 0.57, respectively. The 10-fold cross-validation is also the best result of RF, with R² at the city, county, and township scales being 0.996, 0.87, and 0.88, respectively. The SVM is followed by R² at the city, county, and township scales of 0.95, 0.20, and 0.66, respectively. The MLR is followed by R² at the city, county, and township scales of 0.993, 0.03, and 0.56, respectively. The BPNN has the worst results, with R² at the city, county, and township scales of 0.63, 0.01, and 0.55, respectively. Therefore, the RF was chosen as the best method for the next step of research.

(2) Compared with the results of population simulation based on NPP-VIIRS nighttime light images from 2013 to 2018 and year by year at the city，county and township scale at the Chengdu-Chongqing urban agglomeration, the NPP-VIIRS was selected to use the RF model to map the spatial distribution of the population at a resolution of 500m at the scale of city, county, and township. The R² at calibration of different scales were 0.998, 0.96 and 0.96, and were 0.996, 0.87 and 0.88 in validation. The RF models based on Luojia1-01 nighttime light images were also used as the city, county, and township scale’s population spatial distribution mapping methods to produce the 100m resolution population spatial distribution maps. The R² at calibration of different scales were 0.97, 0.81 and 0.93, and were 0.73, 0.01 and 0.65 in validation. The population spatial distribution maps with 500m resolution from 2013 to 2018 were cross-validated with the LandScan and Worldpop population rasterized products with 1km resolution. The results showed that the accuracy of the population spatial distribution map (R²=0.48) was between LandScan (R²= 0.46) and Worldpop (R²=0.64); The population spatial distribution map and Worldpop population rasterized map product with 100m resolution at 2018 were also cross-validated, and the results showed that the accuracy of the population spatial distribution map (R²=0.38) was lower than Worldpop (R²=0.65).

(3) The characteristics of spatial distribution and laws of temporal changes were studied at the grid scale using the maps of population spatial distribution. The results show that the population of Chengdu-Chongqing urban agglomeration was high in the west and low in the east, with a high in the middle and low in the surrounding area. The population was mainly concentrated in the northwestern part of Chengdu and the central and southern part of Chongqing, and gradually decreased from the densely populated areas to the surrounding area. In the middle of the two densely populated areas, there are moderately dense areas near the Sichuan Basin; The population of the two densely populated areas was growing rapidly, and the population of the denser areas was rapidly decreasing. The sporadic densely populated areas near the Sichuan Basin also showed a rapid growth trend, while the population of the remaining large areas with medium density, sparsely populated areas and extremely sparsely populated areas changed little.

[1] Zeng C, Zhou Y, et al. Population spatialization in China based on night-time imagery and land use data [J]. Int. J. Remote Sens. 2011, 24, 9599–9620.

[2] 胡焕庸. 论中国人口之分布[M]. 北京:科学出版社, 1983, 52-92.

[3] 张车伟, 李庆红. 中国人口合理分布研究:人口空间分布与区域协调发展[M]. 北京:中国社会科学出版社, 2015.

[4] Clark J I, Rhind D W. Population Data and Global Environmental Change [J]. ISSC/Unesco, Paris (HDP Report No. 3). 1992.

[5] Dilley M, Chen R S, et al. Natural Disaster Hotspots: A Global Risk Analysis. Hazard Management Unit, World Bank, Washington DC. 2005.

[6] Garb J L, Cromley R G, et al. Estimating populations at risk for disaster preparedness and response [J]. J. Homel. Secur. Emerg. Manag. 2007, 4 (1) (Article 3).

[7] King G, Bailey G, et al. Active tectonics and human survival strategies [J]. J.Geophys. Res. 1994, 99, 20063-20078.

[8] Rabelo L, Sepulveda J, et al. Disaster prevention and management for the NASA shuttle during lift-off [J]. Disaster Prev. Manag. 2006, 15 (2), 262-274.

[9] Su M D, Lin M C, et al. Multi-layer multi-class dasymetric mapping to estimate population distribution [J]. Sci. Total Environ. 2010, 408 (20), 4807-4816.

[10] Vitousek P, Mooney H A, et al. Human domination of Earth ecosystems [J]. Science. 1997, 277, 494-499.

[11] Obienusi E A, Ekezie M, et al. Investigating the place of population increase and level of income earning in food security a study of yam supply in Awka urban, Anambra state of Nigeria [J]. J. Nat. Sci. Res. 2014, 11, 11–21.

[12] Hutchinson J, Hutchinson J. Population and food supply [J]. BMJ-Br. Med. J. 1969, 3893, 320.

[13] Wang C, Liu Y, et al. Spatiotemporal decoupling between population and construction land in urban and rural Hubei province [J]. Sustainability. 2017, 7, 1258.

[14] Zhang Y. Research on the effects of changes in the age structure of China’s population to energy consumption: Based on the provincial dynamic panel data by GMM [J]. China Popul. Resour. Environ. 2015, 11, 69–74.

[15] Yigzaw W, Hossain F. Water sustainability of large cities in the United States from the perspectives of population increase, anthropogenic activities, and climate change [J]. Earths Future. 2016, 12, 603–617.

[16] Atasoylu G, Evci E D, et al. The household garbage in the western coast region of Turkey and its relationship with the socio-economic characteristics [J]. J. Environ. Biol. 2007, 28, 225–229.

[17] János N. Climate change, agricultural land use and food safety [J]. Econ. Aic. 2014, 7, 149–157.

[18] Li L, Lei Y, et al. Research on sustainable development of resource-based cities based on the DEA approach: A case study of Jiaozuo, China [J]. Math. Probl. Eng. 2016, 3, 1–10.

[19] Miao C L, Sun L Y, et al. The studies of ecological environmental quality assessment in Anhui Province based on ecological footprint [J]. Ecol. Indic. 2016, 60, 879-883.

[20] 张善余. 中国人口地理[M]. 北京:科学出版社, 2003: 400-403.

[21] 乔晓春, 王存同. 2010年人口普查:挑战与对策研讨会综述[J]. 人口与发展, 2010, 16(2): 75-78.

[22] 徐凤娇. 人口数据空间化格网尺度适宜性研究——以安徽省为例[D]. 延边大学, 2017.

[23] 胡云锋, 赵冠华, 张千力. 基于夜间灯光与LUC数据的川渝地区人口空间化研究[J]. 地球信息科学学报, 2018, 20(1): 68-78.

[24] Linard C, Alegana V A, et al. A High Resolution Spatial Population Database of Somalia for Disease Risk Mapping [J]. International Journal of Health Geographics, 2010, 9:45.

[25] 王广州. 大数据时代中国人口科学研究与创新[J]. 人口研究, 2015, 39(5): 15-26.

[26] 李德仁, 余涵若, 李熙. 基于夜光遥感影像的“一带一路”沿线国家城市发展时空格局分析[J]. 武汉大学学报(信息科学版), 2017, 42(6): 711-720.

[27] Langford M, Higgs G, et al. Urban population distribution models and service accessibility estimation [J]. Comput. Environ. Urban Syst. 2008, 32, 66–80.

[28] Zeng C, Zhou Y, et al. Population spatialization in China based on night-time imagery and land use data [J]. Int. J. Remote Sens. 2011, 24, 9599–9620.

[29] Silvan-Cardenas J L, Wang L, et al. Assessing fine-spatial-resolution remote sensing for small-area population estimation [J]. Int. J. Remote Sens. 2010, 31, 5605–5634.

[30] Calka B, Nowak Da Costa J, et al. Fine scale population density data and its application in risk assessment [J]. Geomat. Nat. Hazards Risk 2017, 8, 1440–1455.

[31] Center for International Earth Science Information Network-CIESIN-Columbia University. Gridded Population of the World, Version 4(GPWv4): Population Density, Revision 11[DB/OL].

[32] NASA Socioeconomic Data and Applications Center (SEDAC): Palisades, NY, USA, 2018.

[33] Schiavina M, Freire S, et al. GHS Population Grid, Derived from GPW4, Multitemporal (1975, 1990, 2000, 2015) [DB/OL]; European Commission, Joint Research Centre, JRC Data Catalogue: Ispra, Italy, 2015.

[34] Goldewijk K K, Beusen A, et al. The HYDE 3.1 spatially explicit database of human-induced global land-use change over the past 12,000 years [J]. Glob. Ecol. Biogeogr. 2010, 20, 73–86.

[35] Goldewijk K K, Beusen A, et al. Long-term dynamic modeling of global population and built-up area in a spatially explicit way: HYDE 3.1 [J]. Holocene 2010, 20, 565–573.

[36] Goldewijk K K, Beusen A, et al. Anthropogenic land use estimates for the Holocene–HYDE 3.2 [J]. Earth Syst. Sci. Data 2017, 9, 927–953.

[37] WorldPop (www.worldpop.org-School of Geography and Environmental Science, University of Southampton; Department of Geography and Geosciences, University of Louisville; Departement de Geographie, Universite de Namur); Center for International Earth Science Information Network (CIESIN), Columbia University. WorldPop [DB/OL]; University of Southampton: Southampton, UK, 2018.

[38] Smith A, Bates P D, et al. New estimates of flood exposure in developing countries using high-resolution population data [J]. Nat. Commun. 2019, 10, 1814.

[39] Center for International Earth Science Information, N.C., Columbia University. 2005. “Gridded Population of the World Version 3 (GPWv3), Palisades, NY.”

[40] Bhaduri B, E Bright P, et al. LandScan USA: A High-Resolution Geospatial and Temporal Modeling Approach for Population Distribution and Dynamics [J]. Geojournal. 2007, 69 (1–2): 103–117.

[41] UNEP. Global Population Set to Hit 7 Billion on Halloween [J]. Western Farm Press Exclusive Insight. 2011.

[42] Fu J, D Jiang, et al. 1 Km Grid Population Dataset of China (2005, 2010) [J]. Acta Geographic Sin. 2014, 69: 136-139.

[43] Pesaresi M, Ehrlich D, et al. The global human settlement layer from landsat imagery [J]. In Proceedings of the 2016 IEEE International Geoscience and Remote Sensing Symposium (IGARSS), Beijing, China, 10 July 2016; pp. 7276–7279.

[44] Kibret S, Lautze J, et al. Malaria impact of large dams in sub-Saharan Africa: Maps, estimates and predictions [J]. Malar. J. 2015, 14, 339.

[45] World Resources Institute. World Resources Report 2010-2011: Decision Making in a Changing Climate [J]. Sustainability. 2011, 4, 305.

[46] Smith, K. We are seven billion [J]. Nat. Clim. Chang. 2011, 1, 331–335.

[47] Gleeson T, Wada Y, et al. Water balance of global aquifers revealed by groundwater footprint [J]. Nature. 2012, 488, 197–200.

[48] Balk D L, Nghiem S V, et al. Up and out: A multifaceted approach to characterizing urbanization in Greater Saigon, 2000-2009 [J]. Landsc. Urban Plan. 2019, 187, 199–209.

[49] Melchiorri M, Florczyk A J, et al. Unveiling 25 years of planetary urbanization with remote sensing: Perspectives from the Global Human Settlement Layer [J]. Remote Sens. 2018, 10, 768.

[50] Goldewijk K K. Three centuries of global population growth: A spatial referenced population (density) database for 1700–2000 [J]. Popul. Environ. 2005, 26, 343–367.

[51] Seto K C, Gueneralp B, et al. Global forecasts of urban expansion to 2030 and direct impacts on biodiversity and carbon pools [J]. Proc. Natl. Acad. Sci. USA 2012, 109, 16083–16088.

[52] Thomson D R, Linard C, et al. Extending data for urban health decision n-making: A menu of new and potential neighborhood-level health determinants datasets in LMICs [J]. J. Urban Health-Bull. N. Y. Acad. Med. 2019, 96, 514–536.

[53] Ouma P O, Maina J, et al. Access to emergency hospital care provided by the public sector in sub-Saharan Africa in 2015: A geocoded inventory and spatial analysis [J]. Lancet Glob. Health. 2018, 6, e342–e350.

[54] Sorichetta A, Bird T J, et al. Mapping internal connectivity through human migration in malaria endemic countries [J]. Sci. Data 2016, 3, 160066.

[55] Goldewijk K K, Ramankutty N. Land cover change over the last three centuries due to human activities: The availability of new global data sets [J]. GeoJournal. 2004, 61, 335–344.

[56] Ellis E C, Goldewijk K K, et al. Anthropogenic transformation of the biomes, 1700 to 2000 [J]. Glob. Ecol. Biogeogr. 2010, 19, 589–606.

[57] Houweling S, van der Werf G R, et al. Early anthropogenic CH4 emissions and the variation of CH4 and 13CH4 over the last millennium [J]. Glob Biogeochem Cycles. 2008, 22, GB10021.

[58] Gaston K J, Blackburn T M, et al. Habitat conversion and global avian biodiversity loss [J]. Proc. R. Soc. B-Biol. Sci. 2003, 270, 1293-1300.

[59] Maisels F, Strindberg S, et al. Devastating decline of forest elephants in Central Africa [J]. PLoS ONE. 2013, 8, e59469.

[60] MacPherson P, Khundi M, et al. Disparities in access to diagnosis and care in Blantyre, Malawi, identified through enhanced tuberculosis surveillance and spatial analysis [J]. BMC Med. 2019, 17, 21.

[61] Sun Z C, Xu R, et al. High-resolution urban land mapping in China from Sentinel 1A/2 imagery based on Google Earth Engine [J]. Remote Sens. 2019, 11, 752.

[62] Ye T T, Zhao N Z, et al. Improved population mapping for China using remotely sensed and points-of-interest data within a random forests model [J]. Sci. Total Environ. 2019, 658, 936-946.

[63] Tan M, Lin K, et al. Spatialization of population in the Pearl River Delta in 30m grids using random forest model [J]. Prog. Geogr. 2017, 36, 1304-1312.

[64] Goldewijk K K, Beusen A, et al. The HYDE 3.1 spatially explicit database of human-induced global land-use change over the past 12,000 years [J]. Glob. Ecol. Biogeogr. 2010, 20, 73-86.

[65] Gleeson T, Wada Y, et al. Water balance of global aquifers revealed by groundwater footprint [J]. Nature. 2012, 488, 197-200.

[66] Kibret S, Lautze J, et al. Malaria impact of large dams in sub-Saharan Africa: Maps, estimates and predictions [J]. Malar. J. 2015, 14, 339.

[67] Melchiorri M, Florczyk A J, et al. Unveiling 25 years of planetary urbanization with remote sensing: Perspectives from the Global Human Settlement Layer [J]. Remote Sens. 2018, 10, 768.

[68] Smith A, Bates P D, et al. New estimates of flood exposure in developing countries using high-resolution population data [J]. Nat. Commun. 2019, 10, 1814.

[69] Bai Z Q, Wang J L, et al. Accuracy assessment of multi-source gridded population distribution datasets in China [J]. Sustainability. 2018, 10, 1363.

[70] Leyk S, Uhl J H, et al. Assessing the accuracy of multi-temporal built-up land layers across rural-urban trajectories in the United States [J]. Remote Sens. Environ. 2018, 204, 898–917.

[71] Ye T T, Zhao N Z, et al. Improved population mapping for China using remotely sensed and points-of-interest data within a random forests model [J]. Sci. Total Environ. 2019, 658, 936–946.

[72] Yang X H, Wang N B, et al. Regionalization of population distribution based on spatial analysis [J]. Acta Geogr. Sin. 2002, 57, 76–81.

[73] Sliuzas R, Kuffer M, et al. Assessing the quality of Global Human Settlement Layer products for Kampala, Uganda [J]. In Proceedings of the 2017 Joint Urban Remote Sensing Event, Dubai, UAE, 6–8 March 2017.

[74] Zhu R, Lin D, et al. Geo-tagged social media data-based analytical approach for perceiving impacts of social events [J]. ISPRS Int. J. Geo-Inf. 2019, 8(1), 15.

[75] Atzberger C. Advances in Remote Sensing of Agriculture: Context Description, Existing Operational Monitoring Systems and Major Information Needs [J]. Remote Sens. 2013, 5, 949-981.

[76] Rogan J, Chen D. Remote sensing technology for mapping and monitoring land-cover and land-use change [J]. Prog. Plan. 2004, 61, 301-325.

[77] Li X, Elvidge C, et al. Remote sensing of night-time light [J]. Int. J. Remote Sens. 2017, 38, 5855–5859.

[78] Elvidge C D, Baugh K E, et al. Relation between satellite observed visible-near infrared emissions, population, economic activity and electric power consumption [J]. International Journal of Remote Sensing. 1997, 18(6): 1373-1379.

[79] Levin N, Duke Y. High spatial resolution night-time light images for demographic and socio-economic studies [J]. Remote Sensing Environment. 2012, 119: 1-10.

[80] Colomb R, Alonso C, et al. SAC-C mission and the international am constellation for earth observation [J]. Acta Astronout. 2003, 52: 995-1005.

[81] Letu H, Hara M, et al. A Saturated light correction method for DMSP/OLS nighttime satellite imagery [J]. IEEE Transactions on Geoscience and Remote Sensing. 2012, 50(2): 389-396.

[82] Wu J, Wang Z, et al. Exploring factors affecting the relationship between light consumption and GDP based on DMSP/OLS nighttime satellite imagery [J]. Remote Sensing of Environment. 2013, 134: 111-119.

[83] Henderson J V, Storeyard A, et al. Measuring economic growth from outer space [J]. American economic review. 2012, 102(2): 994-1028.

[84] Zhao N Z, Ghosh T, et al. Mapping spatio-temporal changes of Chinese electric power consumption using night-time imagery [J]. International Journal of Remote Sensing. 2012, 33(20): 6304-6320.

[85] Townsend A C, Bruce D A. The use of night-time lights satellite imagery as a measure of Australia's regional electricity consumption and population distribution [J]. International Journal of Remote Sensing. 2010, 31(16): 4459-4480.

[86] 李翔, 朱江, 尹向东. 基于夜间灯光数据的中国经济发展时空格局特征 [J]. 地球信息科学学报, 2019, 21(3): 417-426.

[87] He C Y, Ma Q, et al. Spatiotemporal Dynamics of Electric Power Consumption in Chinese Mainland from 1995 to 2008 Modeled Using DMSP/OLS Stable Nighttime Lights Data [J]. Journal of Geographical Sciences. 2012, 22(1): 125-136.

[88] Chand T K, Badarinath K, et al. Spatial characterization of electrical power consumption patterns over India using temporal DMSP-OLS night-time satellite data [J]. International Journal of Remote Sensing. 2009, 30(3): 647-661.

[89] Briggs D.J, Gulliver J, et al. Dasymetric modelling of small-area population distribution using land cover and light emissions data [J]. Remote Sens. Environ. 2007, 108, 451–466.

[90] Bagan H, Yamagata Y. Analysis of urban growth and estimating population density using satellite images of nighttime lights and land-use and population data [J]. GISci. Remote Sens. 2015, 52, 765–780.

[91] Amaral S, Monteiro A M V, et al. DMSP/OLS night-time light imagery for urban population estimates in the Brazilian Amazon [J]. Int. J. Remote Sens. 2006, 27, 855–870.

[92] Sun W, Zhang X, et al. Estimating Population Density Using DMSP-OLS Night-Time Imagery and Land Cover Data [J]. IEEE J. Sel. T op. Appl. Earth Obs. Remote Sens. 2017, 10, 2674–2684.

[93] Ziskin D, Baugh K, et al. Methods used for the 2006 radiance lights [J]. Proc. Asia Pac. Adv. Netw. 2010, 30, 131-142.

[94] Li X, Zhou Y. Urban mapping using DMSP/OLS stable night-time light: A review [J]. Int. J. Remote Sens. 2017, 38, 6030–6046.

[95] Elvidge C D, Ziskin D, et al. A Fifteen Year Record of Global Natural Gas Flaring Derived from Satellite Data [J]. Energies 2009, 2, 595–622.

[96] Hsu F C, Baugh K E, et al. DMSP-OLS radiance calibrated nighttime lights time series with intercalibration [J]. Remote Sens. 2015, 7, 1855–1876.

[97] Elvidge C D, Baugh K E, et al. Why VIIRS data are superior to DMSP for mapping nighttime lights [J]. Proc. Asia Pac. Adv. Netw. 2013, 35, 62.

[98] Alahmadi M, Atkinson P M. Three-Fold Urban Expansion in Saudi Arabia from 1992 to 2013 Observed Using Calibrated DMSP-OLS Night-Time Lights Imagery [J]. Remote Sens. 2019, 11, 2266.

[99] Liu A, Wei Y, et al. Estimation of Cargo Handling Capacity of Coastal Ports in China Based on Panel Model and DMSP-OLS Nighttime Light Data [J]. Remote Sens. 2019, 11, 582.

[100] Stevens F R, Gaughan A E, et al. Disaggregating Census Data for Population Mapping Using Random Forests with Remotely-Sensed and Ancillary Data [J]. PLoS ONE. 2015, 10, e0107042.

[101] Chen J, Fan W, et al. Fitting Chinese cities’ population distributions using remote sensing satellite data [J]. Ecol. Indic. 2019, 98, 327–333.

[102] Gaughan A E, Stevens F R, et al. Exploring nationally and regionally defined models for large area population mapping [J]. Int. J. Digit. Earth. 2015, 8, 989–1006.

[103] Chen X, Nordhaus W. A Test of the New VIIRS Lights Data Set: Population and Economic Output in Africa [J]. Remote Sens. 2015, 7, 4937–4947.

[104] Shi K, Huang C, et al. Evaluation of NPP-VIIRS night-time light composite data for extracting built-up urban areas [J]. Remote Sens. Lett. 2014, 5, 358–366.

[105] Sahoo S, Gupta P K, et al. Comparative analysis between VIIRS-DNB and DMSP-OLS night-time light data to estimate electric power consumption in Uttar Pradesh, India [J]. Int. J. Remote Sens. 2019, 41, 2565–2580.

[106] Shi K, Yu B, et al. Evaluating the Ability of NPP-VIIRS Nighttime Light Data to Estimate the Gross Domestic Product and the Electric Power Consumption of China at Multiple Scales: A Comparison with DMSP-OLS Data [J]. Remote Sens. 2014, 6, 1705–1724.

[107] Zhao J, Chen Y, et al. Residential carbon dioxide emissions at the urban scale for county-level cities in China: A comparative study of nighttime light data [J]. J. Clean. Prod. 2018, 180, 198–209.

[108] G Zhang, J Wang, et al. On-orbit geometric calibration and validation of Luojia1-01 night-light satellite [J]. Remote Sensing. 2019, 11(3), 264.

[109] X Li, L Zhao, et al. Mapping urban extent using Luojia1-01 nighttime light imagery [J]. Sensors. 2018, 18(11), 3665.

[110] X Li, X Li, et al. A preliminary investigation of Luojia-1 night-time light imagery [J]. Remote Sensing Letters. 2019, 10(6), 526–535.

[111] Ahola T, Virrantaus K, et al. A spatio-temporal population model to support risk assessment and damage analysis for decision-making [J]. Int. J. Geogr. Inf. Sci. 2007, 21, 935-953.

[112] Aubrecht C, Ozceylan D, et al. Multi-level geospatial modeling of human exposure patterns and vulnerability indicators [J]. Nat. Hazards. 2013, 68, 147-163.

[113] Linard C, Kabaria C W, et al. Modelling changing population distributions: an example of the Kenyan Coast, 1979-2009 [J]. Int. J. Digit. Earth. 2017, 10, 1017-1029.

[114] Dmowska A, Stepinski T F. A high-resolution population grid for the conterminous United States: The 2010 edition [J]. Comput. Environ. Urban. 2017, 61, 13-23.

[115] Tian Y Z, Yue T X, et al. Modeling population density using land cover data [J]. Ecol. Model. 2005, 189, 72-88.

[116] Linard C, Gilbert M, et al. Assessing the use of global land cover data for guiding large area population distribution modelling [J]. Geo Journal. 2011, 76, 525-538.

[117] Huang Q, He C, et al. Detecting the 20year city-size dynamics in China with a rank clock approach and DMSP/OLS nighttime data [J]. Landscape and Urban Planning. 2015, 137, 138-148.

[118] Huang Q X, Yang Y, et al. A simulation study on the urban population of China based on nighttime light data acquired from DMSP/OLS [J]. Sustainability-Basel. 2016, 8, 13.

[119] Lo C P. Modeling the population of China using DMSP operational linescan system nighttime data [J]. Photogramm. Eng. Remote Sens. 2001, 67, 1037-1047.

[120] Zhou Y, Li X, et al. A global record of annual urban dynamics (1992–2013) from nighttime lights [J]. Remote Sens. Environ. 2018, 219, 206-220.

[121] Briggs D J, Gulliver J, et al. Dasymetric modelling of small-area population distribution using land cover and light emissions data [J]. Remote Sens. Environ. 2007, 108, 451-466.

[122] Doll C N H, Pachauri S. Estimating rural populations without access to electricity in developing countries through night-time light satellite imagery [J]. Energy Policy. 2010, 38, 5661-5670.

[123] Li X M, Zhou W Q. Dasymetric mapping of urban population in China based on radiance corrected DMSP-OLS nighttime light and land cover data [J]. Sci. Total Environ. 2018, 643, 1248-1256.

[124] Hu L J, He Z Y, et al. Adaptive multi-scale population spatialization model constrained by multiple factors: A case study of Russia [J]. Cartogr. J. 2017, 54, 265-282.

[125] Stevens F R, Gaughan A E, et al. Disaggregating census data for population mapping using random forests with remotely-sensed and ancillary data [J]. PLoS One. 2015, 10, 22.

[126] Yang X C, Yue W Z, et al. Spatial improvement of human population distribution based on multi-sensor remote-sensing data: an input for exposure assessment [J]. Int. J. Remote Sens. 2013, 34, 5569-5583.

[127] Yao Y, Liu X, et al. Mapping fine-scale population distributions at the building level by integrating multisource geospatial big data [J]. Int. J. Geogr. Inf. Sci. 2017, 31, 1-25.

[128] Blanford J I, Huang Z J, et al. Geo-Located Tweets. Enhancing mobility maps and capturing cross-border movement [J]. PLoS One. 2015, 10, 16.

[129] Patel N N, Stevens F R, et al. Improving large area population mapping using geotweet densities [J]. Trans. GIS. 2017, 21, 317-331.

[130] Zhu R, Lin D, et al. Geo-tagged social media data-based analytical approach for perceiving impacts of social events [J]. ISPRS Int. J. Geo-Inf. 2019, 8(1), 15.

[131] Y. Ju, M. Kim, et al. Detection of malicious code using the direct hashing and pruning and support vector machine[J]. Concurrency Comput., Pract. Exper. 2020, 32(18).

[132] Wei J, Li Z, et al. Improved 1 km resolution PM2.5 estimates across China using enhanced space–time extremely randomized trees[J]. Atmos. Chem. Phys. 2020, 20(6), 3273–3289.

[133] Wei J, Huang W, et al. Estimating 1-km-resolution PM2.5 concentrations across China using the space-time random forest approach. Remote Sensing of Environment[J]. 2019, 231, 111221–.

[134] Chen G, Li S, et al. A machine learning method to estimate PM2.5 concentrations across China with remote sensing, meteorological and land use information[J]. Sci. Total Environ. 2018, 636, 52-60.