基于决策树与NDVI时序变化检测的撂荒耕地的地形特征—以巫山县为例
DOI:
作者:
作者单位:

1.重庆工商大学环境与资源学院;2.重庆工商大学公共管理学院;3.重庆财经学院公共管理学院;4.生态环境空间信息数据挖掘与大数据集成重庆市重点实验室

作者简介:

通讯作者:

中图分类号:

基金项目:

国家社会科学基金项目(22XJY006);重庆市自科基金面上项目(cstc2020jcyj-msxmX0582)


Topographic characteristics of abandoned farmland based on decision tree and NDVI time series change detection—A case study in Wushan County
Author:
Affiliation:

School of Public Administration,Chongqing Technology and Business University

Fund Project:

  • 摘要
  • |
  • 图/表
  • |
  • 访问统计
  • |
  • 参考文献
  • |
  • 相似文献
  • |
  • 引证文献
  • |
  • 资源附件
  • |
  • 文章评论
    摘要:

    [目的]揭示山区撂荒耕地的地形特征,对耕地保护和粮食安全至关重要。[方法]本研究以重庆市巫山县为研究区域,利用谷歌地球引擎(Google Earth Engine, GEE)云平台上的Landsat TM/OLI和Sentinel-2数据,采用决策树与时间序列 NDVI变化检测法,对2017-2021年研究区的撂荒地信息进行提取和分析。[结果]①从时间序列特征上看,研究区内撂荒地面积整体呈上升趋势,2017年到2021年的增加量为2123.50hm2,增长率为19.61%,区间内曲线形态上表现为“W”型特征。空间上撂荒地呈全局分散,局部集中特征,主要沿着水系走向分布,显著集中河流两侧,周边被坡耕地围绕。②研究区内撂荒地在不同的高程带和坡度带分布不同。撂荒地主要集中于高程1000m以下和坡度5°-20°范围内。高程1500m以下的区域,撂荒地面积和撂荒率表现“先增后减”的规律,2019年达到最高点;高程1500m以上的区域,撂荒地面积和撂荒率随着时间的变化呈现“先减后增”的规律,在2020年达到了最低值。③撂荒地在不同地形位等级下的分布指数表现为持续减少型,地形梯度1级、2级的分布指数大于1,为撂荒地的优势区。[结论]决策树与NDVI时序变化检测法结合能够精准识别撂荒地,识别精度为83.59%。本研究成果可为其他相似地形区域的撂荒地提供参考和借鉴,并提供保障国家粮食安全和促进区域可持续发展的相关政策的依据。

    Abstract:

    [Objective] Revealing the terrain characteristics of abandoned farmland in mountainous areas is crucial to farmland protection and food security. [Methods] This study takes Wushan County of Chongqing as the research area,uses Landsat TM/OLI and Sentinel-2 data on the Google Earth Engine (GEE) cloud platform, and adopts decision tree and time series NDVI change detection methods. Extract and analyze abandoned land information in the study area from 2017 to 2021.[Results]①From the perspective of time series characteristics, the area of abandoned land in the study area shows an overall upward trend, the increase from 2017 to 2021 is 2123.50hm2, with a growth rate of 19.61%, and the curve shape within the interval shows a "W"-shaped characteristic. Spatially, abandoned land is globally dispersed and locally concentrated. It is mainly distributed along the direction of the water system, significantly concentrated on both sides of the river, and is surrounded by sloping farmland. ②The abandoned land in the study area is distributed differently in different elevation zones and slope zones. Abandoned land is mainly concentrated in elevations below 1000m and slopes within the range of 5°-20°. In areas with an elevation below 1500m, the area of abandoned land and the rate of abandonment show a pattern of "first increasing and then decreasing", reaching the highest point in 2019; in areas with an elevation above 1500m, the area of abandoned land and the rate of abandonment show a pattern of "first decreasing and then decreasing" over time. The law of "increase" reached the lowest value in 2020.③The distribution index of abandoned land under different terrain levels is a continuous decrease type, and the distribution index of terrain gradient 1 and 2 is greater than 1, which is the dominant area of abandoned land.[Conclusion] The combination of decision tree and NDVI time series change detection method can accurately identify abandoned land, and the recognition accuracy is 83.59%. The results of this study can provide reference for other abandoned land in similar terrain areas.

    参考文献
    相似文献
    引证文献
引用本文
分享
文章指标
  • 点击次数:
  • 下载次数:
  • HTML阅读次数:
  • 引用次数:
历史
  • 收稿日期:2023-12-12
  • 最后修改日期:2024-03-11
  • 录用日期:2024-03-11
  • 在线发布日期:
  • 出版日期: