测绘学报 ›› 2022, Vol. 51 ›› Issue (4): 577-586.doi: 10.11947/j.AGCS.2022.20220086

• 同济大学测绘学科创建90周年 • 上一篇    下一篇

城市典型要素遥感智能监测与模拟推演关键技术

冯永玖1,2, 李鹏朔1,2, 童小华1,2, 席梦镕1,2, 柳思聪1,2, 许雄1,2   

  1. 1. 同济大学测绘与地理信息学院, 上海 200092;
    2. 上海市航天测绘遥感与空间探测重点实验室, 上海 200092
  • 收稿日期:2022-02-14 修回日期:2022-03-23 发布日期:2022-04-24
  • 作者简介:冯永玖(1981-),男,教授,博士生导师,研究方向为SAR与多光谱遥感数据智能处理、地理信息智能建模等。.E-mail:yjfeng@tongji.edu.cn
  • 基金资助:
    国家自然科学基金(42071371);上海市人才发展资金(2021047)

Key technologies for remote sensing intelligent monitoring and simulation of urban spatial elements

FENG Yongjiu1,2, LI Pengshuo1,2, TONG Xiaohua1,2, XI Mengrong1,2, LIU Sicong1,2, XU Xiong1,2   

  1. 1. College of Surveying and Geo-informatics, Tongji University, Shanghai 200092, China;
    2. The Shanghai Key Laboratory of Space Mapping and Remote Sensing for Planetary Exploration, Tongji University, Shanghai 200092, China
  • Received:2022-02-14 Revised:2022-03-23 Published:2022-04-24
  • Supported by:
    The National Natural Science Foundation of China (No. 42071371); The Shanghai Talent Development Fund (No. 2021047)

摘要: 城市典型要素遥感智能监测与模拟推演的理论、方法与应用,对于国土空间规划与管理,城市规划与综合治理,区域决策与管理等均具有关键支撑作用。针对覆盖要素和驱动要素复杂非线性,本文研发了协同多源遥感数据的智能识别方法,实现了精细化高可信覆盖要素分类;协同遥感、POI兴趣点和时空大数据等多源数据,有效探测和识别了要素变动的驱动力。在此基础上,开展了空间演变机理挖掘、空间统计建模、启发式智能建模,并应用于土地利用、城市扩张、生态演变、碳储量等。同时,研发了聚焦城市生长推演的UrbanCA平台以及聚焦多类土地利用变化推演的Futureland平台,集成了自主研发的模拟推演系列方法并以长三角为主要区域进行了验证。

关键词: 城市空间要素, 遥感监测, 智能建模, 模拟推演, 情景预测

Abstract: For various urban spatial elements, the method development and practical applications are in the center of the intelligent monitoring and spatial deduction simulation using multi-source remote sensing and GIS. The monitoring and simulation are of great significance to territorial and spatial planning and management, urban planning and comprehensive control, and regional decision-making and management. The coverage and driving elements in urban areas are complex and nonlinear, thus we have developed a few intelligent identification methods (e.g. the intelligent adaptive decision tree classifier) that use multi-source remote sensing data and can derive highly accurate and reliable coverage element results. By integrating multi-source remote sensing, POI, and spatiotemporal big data, we have developed new methods that can effectively detect and identify the driving forces of urban element changes. Urban simulation and deduction are advanced modeling based on the spatial monitoring of remote sensing for urban management and decision-making. We systematically have studied the urban deduction and prediction method based on urban spatial evolution mechanisms, spatial statistical modeling, and heuristic intelligent modeling, and applied these methods to simulate complex land use, urban expansion, ecological evolution, and carbon storage. Among the platforms available, we have developed two state-of-art software packages (i.e. UrbanCA and Futureland) in which the former focuses on urban growth and the latter focuses on multiple types of land-use change, and both integrate a variety of advanced methods, which have been successfully verified in the Yangtze River Delta.

Key words: urban spatial elements, remote sensing monitoring, intelligent modeling, simulation and deduction, scenario prediction

中图分类号: