地球信息科学学报 ›› 2017, Vol. 19 ›› Issue (5): 653-661.doi: 10.3724/SP.J.1047.2017.00653

• 地理空间分析综合应用 • 上一篇    下一篇

夜间灯光数据驱动的成渝城市群空间形成过程重建及分析

杨任飞1(), 罗红霞1,*(), 周盛2, 程玉丝1, 陈婧祎1, 向海燕1, 雷茜1   

  1. 1. 西南大学 地理科学学院,重庆 400715
    2. 安徽大学 资源与环境工程学院,合肥 230601
  • 收稿日期:2016-08-17 修回日期:2016-12-05 出版日期:2017-05-20 发布日期:2017-05-20
  • 通讯作者: 罗红霞 E-mail:yrf@email.swu.edu.cn;tam_7236@swu.edu.cn
  • 作者简介:

    作者简介:杨任飞(1993- ),男,四川泸州人,硕士生,主要从事遥感与GIS应用研究。E-mail:yrf@email.swu.edu.cn

  • 基金资助:
    中央高校基本科研业务费专项资金资助“耦合夜间灯光数据与元胞自动机的城市扩张模拟研究—以成渝城市群为例”(XDJK2017D025);国家自然科学基金项目“基于遥感—过程耦合模型的岩溶石漠化植被恢复区碳收支及其时空格局特征研究”(41201436)

Restoring and Analyzing the Space Forming Process of Chengdu-Chongqing Urban Agglomeration by Using DMSP/OLS Night-time Light Data

YANG Renfei1(), LUO Hongxia1,*(), ZHOU Sheng2, CHENG Yusi1, CHEN Jingyi1, XIANG Haiyan1, LEI Xi1   

  1. 1. School of Geographic Science, Southwest University, Chongqing 400715, China
    2. School of Resources and Environmental Engineering, Anhui University, Hefei 230601, China.
  • Received:2016-08-17 Revised:2016-12-05 Online:2017-05-20 Published:2017-05-20
  • Contact: LUO Hongxia E-mail:yrf@email.swu.edu.cn;tam_7236@swu.edu.cn

摘要:

2016年4月发布的《成渝城市群发展规划》首次正式确定了成渝城市群的内涵和具体边界,重建成渝城市群的形成过程,有利于把握未来发展趋势,并合理优化与调整其发展过程。在重建技术方面,对DMSP/OLS夜间灯光数据传统的不变目标区域校正法加以改进,将成渝城市群2013年城市市区范围内的全部像元加入校正模型的拟合中,设计了统计数据的校正规则,再通过二分比较法较好地恢复了成渝城市群内各城市建成区的时序空间信息。提取面积与统计面积总体平均相对误差为-0.38%,利用高分辨率Google Earth图像验证的建成区提取准确率达到98.29%,相比其他研究结果,经方法改进后的提取结果精度高且稳定。在结果分析方面,基于提取结果展开对城市群建成区重心转移过程与城市聚合过程的深层次研究,剖析了城市群的内部格局与时空变化特征。分析表明,成渝城市群的聚合情况与《成渝城市群发展规划》高度吻合,城市群已进入快速发育阶段,随着区域差异的持续扩大,成都、重庆都市圈的核心地位逐渐形成,而重庆的发展态势稍好。

关键词: DMSP/OLS, 建成区, 重心, 分离度

Abstract:

In April 2016, Chengdu-Chongqing Urban Agglomeration Development Plan firstly officially confirmed the connotation and specific boundary of Chengdu-Chongqing urban agglomeration. It's conducive to optimize and adjust the trend of development in future and help understanding the spatial patterns and dynamic changes of Chengdu-Chongqing urban agglomeration. Firstly, during the pretreatment process of DMSP/OLS night-time light data, the traditional methods calibrated images by a representative invariant region and those frequently selected regions were far away from the study area, causing uncertainties and errors. We calibrated all pixels of the DMSP/OLS night-time light data within the boundaries of urban-district in 2013. Next, we designed a rule to calibrate the statistical data. On the basis of these two kinds of data, we calculated out the best thresholds of the urban built-up areas in 5 stages of Chengdu-Chongqing urban agglomeration from 1997 to 2013. Comparing extractive areas with statistical areas, the total average relative error is only -0.38%. The result was more stable and accurate than other methods. We tested the extracted built-up polygons (2013) with the certain built-up polygons that sketched on Google Earth historical images. The results showed that the accuracy of extraction had reached 98.29%. Furthermore, we selected the barycenter index and landscape isolation index to analyze the extraction results. We studied the shifting processes of barycenter and the integrating processes of cities and found that: (1) the level of integration was highly consistent with planning and Chengdu-Chongqing urban agglomeration had entered a stage of rapid development; (2) Chongqing metropolitan and Chengdu metropolitan gradually formed and the former achieved greater developments and it also deteriorated regional imbalance. Overall, our research results can be used to optimize the development of Chengdu-Chongqing urban agglomeration, and we provided a template for the study on the space forming processes of urban agglomeration.

Key words: DMSP/OLS, built-up area, barycenter, landscape isolation