地理科学 ›› 2016, Vol. 36 ›› Issue (11): 1743-1750.doi: 10.13249/j.cnki.sgs.2016.11.018

• • 上一篇    下一篇

面向地震应急准备的居民地遥感提取及量化分析

李金香1(), 李亚芳1, 李帅1, 王伟1, 陈勇   

  1. 1.新疆维吾尔自治区地震局, 新疆 乌鲁木齐 830011
  • 收稿日期:2016-02-23 修回日期:2016-05-10 出版日期:2016-11-10 发布日期:2016-11-10
  • 作者简介:

    作者简介:李金香(1984-),女,山东烟台人,工程师,硕士,现从事地震应急及遥感应用研究。E-mail:ljxhappy365@163.com

  • 基金资助:
    中国地震局地震应急青年重点任务(CEA_EDEM-201511)、新疆地震科学基金项目(201510)资助

Remote Sensing Extraction and Quantitative Analysis of Residential Area for Earthquake Emergency Preparedness

Jinxiang Li1(), Yafang Li1, Shuai Li1, Wei Wang1, Yong Chen1   

  1. 1.Earthquake Administration of Xinjiang Uyghur Autonomous Region, Urumqi 830011, Xinjiang, China
  • Received:2016-02-23 Revised:2016-05-10 Online:2016-11-10 Published:2016-11-10
  • Supported by:
    EarthQuake Emergency Youth Key Project of China Earthquake Administration(CEA_EDEM-201511),Earthquake Science Fund Project in Xinjiang(201510).

摘要:

运用灰度共生矩阵、数学形态学等方法提取新疆新源地区高分一号2 m分辨率影像居民地信息,运用目视解译、影像叠加分析、缓冲区分析等方法,进行居民地量化分级,为地震应急准备提供数据支持。结果表明:研究区在地震烈度为度及以下区域,埋压主要集中在单层结构为主的建筑区;当地震烈度高于度且造成多层建筑大面积倒塌时,县城等人口密集区为首要救援区;在地震应急准备时,应对交通条件三等区和交通条件四等区重点关注,增加应急物资储备点,对山区居民地,应考虑道路毁坏情况,转换救援方式,做好应急预案。

关键词: 地震应急, 遥感, 高分一号, 居民地

Abstract:

Residential area is important hazard-bearing body of earthquake disasters. Accurate grasp of the spatial distribution of residential area is an important basis for understanding the earthquake disaster and carrying out the earthquake emergency preparedness. Residential area are changing faster and faster in recent years, the development of remote sensing technology provides advanced means for acquisition spatial information of residential area. The spatial distribution of the real residential area extracted by remote sensing can provide a new data for earthquake emergency preparedness. At the same time, when large destructive earthquake happens, we can rapid determine buried areas and rescue mode according to residential area of quantitative classification result, which has a certain guiding significance on the rescue and evacuation plans. In this article, we use gray level co-occurrence matrix and mathematical morphology methods to extract the spatial distribution of residential area from the 2 m resolution GF-1 satellite remote sensing data, and use visual interpretation, image analysis, buffer analysis to carry out residential area quantitative classification, which can provide data support for the earthquake emergency preparedness. Due to the different earthquake intensity, population distribution and the different types of building structures, the number of buried person in different regions is also different. The residential areas are divided into multi-storey residential areas and bungalow residential areas using visual interpretation, the number of houses is interpretated according to the image characteristic as residential areas attribute data, then the residential areas of buried person distribution are graded through analysis. At the same time, in traffic as the research object, different grade of roads has different road accessibility. We do buffer analysis with different effects ranges for state roads, provincial roads, county roads, township roads and special roads in study area, then the residential areas of rescue convenience degree are graded according to different regional values. The residential areas are divided into four grades, such as first level traffic conditions residential areas, second level traffic conditions residential areas, third traffic conditions residential areas and fourth level traffic conditions residential areas. The results revealed that: Using gray level co-occurrence matrix and mathematical morphology methods can better extract the residential area information of the high resolution GF-1 2 m image. The algorithm of this article has high accuracy and good robustness. However, to ensure data accuracy, extraction results and images were compared and analyzed, and residential areas were extracted semi automatically by the artificial intervention. The quantitative analysis of the residential area revealed that in seismic intensity VIII and the following area, buried zone mainly concentrated in the bungalow areas; When the seismic intensity is higher than VIII degrees and caused large area multi-storey buildings collapsed, densely populated areas of the county are the main rescue areas; In the earthquake emergency preparedness, we should increase the reserve point of emergency supplies especially in the third and fourth level traffic conditions residential areas. At the same time, the mountain residential areas should be considered to converse the rescue methods and do a good job in the emergency plan in the case of road damage.

Key words: earthquake emergency, remote sensing, GF-1, residential land, quantification

中图分类号: 

  • P237