摘要
夜间灯光数据一定程度上表征了人们夜间的活动强度,能综合地反映城市发展情况,基于夜间灯光数据的城市扩展研究已成为热点科学问题。本研究利用DMSP/OLS夜间灯光数据,以干旱区城市乌鲁木齐为研究靶区,借鉴突变检测法(Muta⁃tion Detection),确定了2001—2013年的建成区提取灯光阈值、城市建成区域以及城市重心,对城市扩展进行时间和空间尺度上的分析,并以总体分类精度(Overall Accuracy)和Kappa系数作为评价指标,最终对突变检测法的建成区提取效果进行评价。研究结果表明:①突变检测法下有时可能存在多个分割阈值,与实际建成区对比方能确定最佳分割阈值,且不同年份的最佳分割阈值不具有通用性。②乌鲁木齐市城区面积在2001—2013年整体上呈现出递增趋势,平均增长速度为22.92 km^(2)/a。③2001—2013年期间,乌鲁木齐市城市重心表现为西北-西南-东北方向转移,但均分布在新市区范围内,总体向西北方向转移。利用突变检测法的乌鲁木齐建成区提取总体分类精度平均值达到了81.07%,同时预测结果与统计数据也表现出较好的一致性。综上,主要探讨了乌鲁木齐市建成区的时空格局变化,可为今后进一步拓展研究干旱区城市发展提供重要的科学依据和理论方法。
Night-light data,to some extent,representspeople’s activity intensityand can comprehensively reflect the situation of urban development.Urban expansion research based on night-light data has become a hot scientific issue.In this study,taking Urumqi,an arid region,as the research target,we used DMSP/OLS night-light data to determine the threshold of light extraction,urban built-up area and urban center of gravity in built-up area from 2001 to 2013,and analyzed urban expansionon time and space scales.We used overall accuracy and Kappa coefficient as evaluation indexes to evaluate the extraction effect of built-up area by mutation detection method.Theresults show that①sometimes,there may be multiple segmentation thresholds under the mutation detection method,and the optimal segmentation threshold can be determined only by comparing with the actual built-up area,and the optimal segmentation threshold of different years is not universal.②The urban area of Urumqi showed an increasing trend from 2001 to 2013,with an average growth rate of 22.92 km^(2)/a.③From 2001 to 2013,the urban center of Urumqi shifted from northwest to southwest to northeast,but all of them were distributed within the new urban area,and generally shifted to northwest.④The average accuracy of total classification of Urumqi built-up area extracted by mutation detection method is 81.07%,and the prediction results are consistent with the statistical data.To sum up,we mainly discussedthe spatio-temporal pattern changes of Urumqi built-up areas,which couldprovide important scientific basis and theoretical methods for further research on urban development in arid areas.
作者
贺鑫
张飞
韩奇笑
周婷
刘雪宁
余欣茹
HE Xin;ZHANG Fei;HAN Qixiao;ZHOU Ting;LIU Xuening;YU Xinru(College of Geography and Remote Sensing Sciences,Xinjiang University,Urumqi 830017,China;Key Laboratory of Oasis Ecology,Xinjiang University,Urumqi 830046,China;Key Laboratory of Smart City and Environmental Modeling of Higher Education Institute,Xinjiang University,Urumqi 830017,China)
出处
《地理空间信息》
2023年第8期9-13,共5页
Geospatial Information
基金
国家大学生创新训练项目(XJU-SRT-21021)
中国科学院战略性先导科技专项资助项目(XDA20040400)。