地理学报 ›› 2018, Vol. 73 ›› Issue (9): 1658-1673.doi: 10.11821/dlxb201809004

• 土地利用 • 上一篇    下一篇

撂荒耕地的提取与分析——以山东省庆云县和无棣县为例

肖国峰1,2,3,4(),朱秀芳2,3,4(),侯陈瑶4,夏兴生4   

  1. 1. 北京师范大学环境遥感与数字城市北京市重点实验室,北京100875
    2. 北京师范大学地表过程与资源生态国家重点实验室,北京 100875
    3. 北京市陆表遥感数据产品工程技术研究中心,北京 100875
    4. 北京师范大学地理科学学部遥感科学与工程研究院,北京 100875
  • 收稿日期:2018-03-05 出版日期:2018-09-25 发布日期:2018-09-19
  • 基金资助:
    国家“高分辨率对地观测系统”重大专项;国家自然科学基金青年基金项目(41401479)

Extraction and analysis of abandoned farmland: A case study of Qingyun and Wudi counties in Shandong Province

XIAO Guofeng1,2,3,4(),ZHU Xiufang2,3,4(),HOU Chenyao4,XIA Xingsheng4   

  1. 1. Beijing Key Laboratory for Remote Sensing of Environment and Digital Cities, Beijing Normal University, Beijing 100875, China
    2. State Key Laboratory of Earth Surface Processes and Resource Ecology, Beijing Normal University, Beijing 100875, China
    3. Beijing Engineering Research Center for Global Land Remote Sensing Products, Institute of Remote Sensing Science and Engineering, Beijing Normal University, Beijing 100875, China
    4. Institute of Remote Sensing Science and Engineering, Faculty of Geographical Science, Beijing Normal University, Beijing 100875, China
  • Received:2018-03-05 Online:2018-09-25 Published:2018-09-19
  • Supported by:
    Major Project of High-Resolution Earth Observation System;National Natural Science Foundation for Distinguished Young Scholars of China, No.41401479

摘要:

由于城镇化的快速发展和农村劳动力不断流失,中国部分地区出现大面积的撂荒现象。利用遥感技术可以间接提取撂荒耕地的规模和数量,对耕地的保护和粮食安全有重要意义。以山东省庆云县和无棣县为研究区,基于Landsat数据和HJ1A数据,采用CART决策树分类方法,制作了1990-2017年的土地利用图,制定了撂荒地的识别规则,在此基础之上提取了研究区的撂荒地空间分布、持续撂荒时间分布和撂荒地复垦区域。结果显示:1990-2017年基准期影像的CART决策树分类精度高于85%;1992-2017年间,研究区撂荒地面积最大值为5503.86 hm2,最大撂荒率为5.37%,其中1996-1998年撂荒率最高,2006-2017年撂荒地面积的整体趋势逐年降低;1992-2017年间最大持续撂荒时间为15年,大部分耕地持续撂荒时间在4年之内,少数耕地持续撂荒时间超过10年;1993-2017年撂荒耕地复垦面积最大为2022.3 hm2,最小复垦面积约为20 hm2,复垦率最大值为67.44%,年均复垦率为31.83%。研究结果不仅能够为研究区撂荒驱动因素分析提供数据支撑,而且也可以为其他地区的撂荒耕地识别提供参考。

关键词: CART, 撂荒, 复垦, 耕地, 山东省, Landsat数据

Abstract:

With the rapid development of urbanization and the continuous loss of the rural labor force, large areas of farmland have been abandoned in some regions of China. Remote sensing technology can indirectly help to detect the size and quantity of abandoned farmland, which is of great significance for the protection of farmland and food security. Using Qingyun and Wudi counties of Shandong Province as the study area, and based on the Landsat data and HJ-1A data, this paper uses the CART decision tree classification method to develop land use maps from 1990 to 2017, set down rules to identify abandoned farmland, and explore the spatial distribution, duration, and reclamation area of abandoned farmland. The results showed that the accuracy of the CART decision tree classification was higher than 85% from 1990 to 2017. The maximum abandoned farmland area was 5503.86 hm2 from 1992 to 2017, and the maximum abandonment rate was 5.37%. The rate of farmland abandonment reached the peak from 1996 to 1998, and the overall trend of abandonment has decreased year by year after 2006. The maximum duration of abandonment was 15 years during the period of 1992 to 2017. Most of the durations were within four years and a few durations were more than ten years. From 1993 to 2017, the maximum reclamation area of abandoned farmland was 2022.3 hm2, and the minimum reclamation area was about 20 hm2. The maximum reclamation rate was 67.44% and the annual average reclamation rate was 31.83%. The results from this paper not only help analyze the driving forces of farmland abandonment in the study area, but also provide references for the identification of abandoned farmland in other areas.

Key words: CART, abandonment, reclamation, farmland, Shandong Province, Landsat data