img

官方微信

高级检索

中国沙漠 ›› 2023, Vol. 43 ›› Issue (3): 230-242.DOI: 10.7522/j.issn.1000-694X.2022.00129

• • 上一篇    

基于ChinaCover数据集的绿洲结构数据制图——以河西内陆河流域为例

李森(), 颜长珍   

  1. 中国科学院西北生态环境资源研究院 沙漠与沙漠化重点实验室,甘肃 兰州 730000
  • 收稿日期:2022-09-02 修回日期:2022-10-24 出版日期:2023-05-20 发布日期:2023-05-31
  • 作者简介:李森(1985—),男,山东潍坊人,高级工程师,研究方向为沙漠化/绿洲化监测。E-mail: lisen@lzb.ac.cn
  • 基金资助:
    国家自然科学基金项目(42007430);第二次青藏高原综合科学考察研究项目(2019QZKK0608);甘肃省自然科学基金项目(22JR5RA077);中国科学院沙漠与沙漠化重点实验室开放基金项目(KLDD-2021-001)

Mapping oasis structure data based on ChinaCover dataset: a case study of Hexi Inland River Basin

Sen Li(), Changzhen Yan   

  1. Key Laboratory of Desert and Desertification,Northwest Institute of Eco-Environment and Resources,Chinese Academy of Sciences,Lanzhou 730000,China
  • Received:2022-09-02 Revised:2022-10-24 Online:2023-05-20 Published:2023-05-31

摘要:

生产兼具绿洲独特性和结构特点的绿洲专题数据是目前绿洲学研究亟待解决的关键基础性问题。以河西内陆河流域为研究区,在构建绿洲生态系统的二级分类体系基础上,结合ChinaCover数据和Landsat NDVI影像,利用面向对象随机森林分类和叠加分析方法,完成1975—2020年7期河西内陆河流域绿洲土地结构数据的制图。结果表明:(1)基于ChinaCover数据集的绿洲结构数据制图方法可以快速准确地生产出绿洲专题数据;(2)绿洲生态系统一级类型总体精度为96.17%,二级类型精度为82.64%~92.99%,制图精度明显高于现有土地覆被产品;(3)2020年河西内陆河流域绿洲占研究区比例为8.97%,一级类型以人工绿洲为主,二级类型以耕地和草地为主;(4)近45年河西内陆河流域人工绿洲呈持续扩张趋势,荒漠-绿洲过渡带呈萎缩趋势,自然绿洲呈先萎缩后扩张趋势;人工绿洲的扩张主要是耕地和建设用地的增加,而过渡带和自然绿洲的萎缩主要是草地减少。

关键词: 绿洲制图, 分类体系, 绿洲变化, 精度评价

Abstract:

The production of oasis thematic data with both oasis uniqueness and structural characteristics is a key basic problem to be solved in oasis research at present. In this paper, a secondary classification system of oasis ecosystem was constructed according to the characteristics of oasis. Then, taking the Hexi inland river basin (HIRB) as the study area, the object-oriented random forest classification and superposition analysis method was used to complete the land structure data mapping of oases in the HIRB from 1975 to 2020 in seven periods, combining with ChinaCover data and Landsat NDVI images. The results show that: (1) the oasis structure data mapping method based on ChinaCover data set can produce oasis thematic data quickly and accurately. (2) the overall accuracy of the primary type of the oasis ecosystem is 96.17%, and the accuracy of the secondary type is 82.64%-92.99%, which is significantly higher than that of the existing land cover products. (3) the proportion of oases in HIRB accounted for 8.97% of the study area in 2020, the primary type was mainly artificial oasis, and the secondary type was mainly cultivated land and grassland. (4) the artificial oasis in the HIRB showed a continuous expansion trend, the desert-oasis ecotone showed a shrinking trend, and the natural oasis shrank at first and then expanded in the past 45 years. The expansion of artificial oasis is mainly due to the increase of cultivated land and construction land, while the shrinkage of ecotone and natural oasis is mainly due to the decrease of grassland. This study can provide a simple and accurate data method for oasis mapping in arid areas.

Key words: mapping oasis, classification system, oasis change, accuracy evaluation

中图分类号: