The Research On Urban Nighttime Light Pattern And Its Impacts On Bird Habitats Based On Nighttime Light Data

Urban are regions where humans are intensively concentrated.The sustainable development of the urban area not only depend on the prosperity of social economy,but also are constrained by the ecological environment.The process of urbanization in China is unprecedented and it will definitely continue.The timely acquisition and analysis of urban extent and shape is of great significance in urban planning and management.However,the artificial light at night(ALAN)pattern inside the urban boundary(UB)rarely considered the different ALAN demands for different land use functional zones.The appropriate ALAN intensity can not only save the energy,but also are beneficial to the urban ecosystem.Obtaining the ALAN pattern inside the UB and analyzing the corresponding ecological impacts on bird habitats is of critical significance in optimizing the ALAN pattern,saving energy and protecting the urban environment.In this study,the Hangzhou city was selected as the study area,nighttime light(NTL)data,other types of remote sensing(RS)data and Tencent's Easygo big data were selected as data sources.The research of the extraction of UB,the ALAN supplydemand pattern and its impacts on bird habitats were conducted.The main research contents and conclusions were summarized as follows:(1)The extraction of UBs based on NTL data and Landsat dataCombining the advantages of both NTL data and traditional optical RS data,a fast and effective method for extracting UB was proposed.This method not only solved the discontinuity of the UB extracted from the traditional optical RS data,but also enhanced the accuracy of UB extracted from the NTL data.The VANUI index were firstly constructed,and then based on the proposed concentric zones model(CZM)and the buffer area,the rough urban extent was identified.The modified Canny edge detection method was adopted to obtain the UB.The proposed CZM search mode make the abrupt change points to rest in the urban fringe.This makes the search processes more convenient compared with previous studies.UBs extracted by the proposed method showed a high accordance with ground truth data,with an overall accuracy of 93.87%,it indicated that the accuracy of our method outperformed the traditional methods.In urban fringe with complex land cover types,the modified edge detection method solved the underestimation and overestimation of urban area from the NTL data and enhanced the accuracy of UB.(2)Spatiotemporal dynamics of UBs in Hangzhou cityBased on the method above,the spatiotemporal dynamics of the UBs of Hangzhou were explored through expansion intensity index,spatial direction analysis and the ALAN gravity model.The extent of urban area in Hangzhou increased rapidly from 2000 to 2018,the urban area increased rapidly from 2000 to 2012 with an average annual growth rate more than 14%.The average annual growth rate dropped sharply after 2012.The main expansion directions are east,northeast and north.The average intensity of ALAN inside the UB increased and the centers of gravity of ALAN move from the West Lake to the Qianjiang new city.(3)The ALAN pattern in different land use functional zonesThe Luojia1-01 NTL data were selected as the data source,the Exploratory Spatial Data Analysis and the evaluation indicators of ALAN intensity were used to analyze the ALAN intensity pattern of different land uses functional zones in block scale inside the UB.The Tencent's big data(population density)were used to represent the ALAN demand.The bivariate spatial clustering algorithm and the sliding window normalized cross-correlation(NCC)was introduced to reveal the ALAN supply-demand pattern and its dynamic changes of different land use functional zones in block scale.The hot spots areas were found in regions with intensive human activities.Except for the urban green space(UGS),the overall brightness of other types of land use functional zones were reasonable.The lit regions were observed in some nondevelopment land,they mainly located inside the West Lake Scenic Spot,Xixi wetland and large urban parks.The total area of L-H and H-L areas accounts for 46% of the urban area.The mismatch of ALAN supply-demand pattern inside UGS and nondevelopment land is more serious which dominated by the H-L types.In different time period,the mismatch of ALAN supply-demand inside the non-development land and industrial land become more serious from 18:00 to 22:00.From the perspective of spatial matching degree of ALAN supply-demand,the area of mismatched ALAN supply-demand inside H-H and L-L accounts for 18.67% of the urban area.(4)The impacts of urban ALAN on the bird habitatsBirds are the most representative wild animal in the urban environment and they are highly susceptible to the ALAN.The dominant urban birds in Hangzhou were selected as the research object.First,the fine-scale ALAN pattern was derived from Jilin1-03 B NTL data.Then,the core habitat nodes(CHNs)representing the main habitats for urban birds to inhabit were identified from the land cover map.The high probability corridors(HPCs),indicating high connectivity paths,were derived from Circuitscape model.Finally,the impacts of ALAN patterns on the CHNs and HPCs were analyzed through spatial analyses.A mismatch index(MI)was proposed to evaluate the trade-off between human activities ALAN supply and demand.115 woodland patches covering 13.83% of the study area were selected as CHNs.Most of the CHNs were large urban parks and scenic spots;2923 HPCs covering 3.93% of the study area was identified.The HPCs were small remaining pocket parks and vegetated corridors along the major transport arteries.The ALAN patterns between CHNs and HPCs were obviously different.The lit regions in the CHNs were clustered in a few regions that suffered from concentrated and intensive ALAN.Most of the HPCs were suffered from the diffused light from traffic lanes,the average intensity was relatively lower.Most of the influenced CHNs and HPCs lacked human activity at night,but suffered from ALAN.The regions with high MI values were mainly located inside the CHNs,and the regions with medium and low MI values were mainly HPCs influenced by ALAN.