玛纳斯河流域, 土地利用变化, CA-Markov, 时间间隔," /> 玛纳斯河流域, 土地利用变化, CA-Markov, 时间间隔,"/> Manas River Basin, land use and cover change, CA-Markov, time span,"/> <span>基于<span>CA-Markov</span>模型的多时间跨度土地利用变化模拟<span></span></span>
收藏设为首页 广告服务联系我们在线留言

干旱区地理 ›› 2019, Vol. 42 ›› Issue (6): 1415-1426.

• 地球信息科学 • 上一篇    下一篇

基于CA-Markov模型的多时间跨度土地利用变化模拟

 

靳含1,2,杨爱民1,2,夏鑫鑫1,2,朱磊1,2,张青青1   

  1. 1 新疆农业大学草业与环境科学学院,新疆 乌鲁木齐 830052 2 新疆土壤与植物生态过程实验室,新疆 乌鲁木齐 830052
  • 收稿日期:2019-03-12 修回日期:2019-07-24 出版日期:2019-11-15 发布日期:2019-11-18
  • 通讯作者: 朱磊(1982-),男,新疆库尔勒人,副教授,博士,研究方向为土地利用变化、地理信息系统和遥感应用
  • 作者简介:靳含(1993-),女,陕西咸阳人,硕士研究生,主要研究土地利用变化. E-mail: hanjin199308@163.com
  • 基金资助:
    国家自然科学基金项目(41561019);国家自然科学基金项目(41561103)

Simulation of land use change at different time spans based on CA-Markov model

 

JIN Han1,2,YANG Ai-min1,2,XIA Xin-xin1,2,ZHU Lei1,2,ZHANG Qing-qing1   

  1. 1 College of Pratacultural and Environmental Sciences,Xinjiang Agricultural University,Urumqi  830052,Xinjiang,China; 2 Xinjiang Key Laboratory of Soil and Plant Ecological Processes,Urumqi 830052,Xinjiang,China
  • Received:2019-03-12 Revised:2019-07-24 Online:2019-11-15 Published:2019-11-18

摘要:  

土地利用/覆被变化(Land Use and Cover Change LUCC)模拟是LUCC研究的主要内容和重要手段。时间间隔是模拟过程中的重要参数,对模拟结果精度有何影响,有待深入研究。以新疆玛纳斯河流域典型绿洲区四道河子镇为例,基于遥感影像提取197519851995200020052010年和2015年的土地利用数据,分别以20 a15 a10 a5 a为时间间隔构建CA-Markov模型,模拟2015年土地利用结构,定量探讨时间间隔对CA-Markov模型精度的影响。结果表明:19752015年,四道河子镇LUCC以耕地和草地为主,期间耕地、建设用地迅速扩张,林地、草地和未利用地大幅减少,水域在19852000年呈现小幅增长。耕地的增加和草地及林地的减少是研究区近40 a LUCC最显著的特征。对比模拟结果与实际结果,时间间隔为20 a15 a10 a5 aTFOM分别为70.35%69.18%76.32%88.00%。基于20052010年转移概率的模拟结果更接近于2015年实际结果,适合模拟四道河子镇未来的土地利用变化。土地利用模拟应依据区域LUCC特征确定最佳的时间间隔,提高模拟精度。

关键词: font-size:10.5pt, 玛纳斯河流域')">">玛纳斯河流域, font-size:10.5pt, "> font-size:10.5pt, 土地利用变化')">">土地利用变化, font-size:10.5pt, CA"> CAfont-size:10.5pt, -">-font-size:10.5pt, Markov">Markovfont-size:10.5pt, ')">">, font-size:10.5pt, "> font-size:10.5pt, 时间间隔')">">时间间隔

Abstract:  

The study of Land Use/Cover Change (LUCC) and its dynamics is crucial for human existence and sustainable development. LUCC simulation is the main content and effective method in LUCC research and time span is one of the most important parameters in the simulation process, and thus its effect on the simulation accuracy needs to be further studied. In order to explore LUCC simulation accuracy at different time spans, we took Sidaohezi Town of Manas River Basin in Xinjiang, China as the study area. Seven land use maps were created from remote sensing images (MSS images in the 1975, M images in 1985 and 1995, ETM images in 2000, 2005 and 2010, and OLI images in 2015) and the land surfaces were classified into six land cover categories: cultivated land, woodland, grassland, water area, construction land and unused land. Then the land use transition probability during four different periods (19751995,19852000,19952005 and 20052010) were used to establish the CA-rkov model to simulate land use in 2015 using long-term (20-ears),medium-term (15-and 10-years) and short-term (5-years) intervals respectively. he simulation process was completed by the CA-arkov module in IDRISI software. Finally, our simulated land use maps for 2015 (R1, R2, R3, R4) were compared with the real land use map in 2015 (R), and the Total Figure of Merit (TFOM) indexes were used to evaluate the model's accuracy. he results showed that from 1975 to 2015,cultivated land and grassland were the dominant land use types in Sidaohezi Town. During the study period, cultivated land and construction land expanded rapidly, whereas forest land, grassland and unused land have been greatly reduced, and the water area has experienced a slight increase from 1985 to 2000.Throughout the study period, he increase of cultivated land and decrease of grassland and woodland were the most significant features of LUCC. compared the simulated results with real land use map in 2015,five components of agreement and disagreement were detected as S1,S2,S3,S4 and S5The TFOM indices of 20-, 15-, 10- and 5-years intervals were 70.35%,69.18%,76.32% and 88.00% respectively. It was indicated that smaller time intervals produced higher accuracy. The time span of 5-years was the optimum time interval to simulate the land use change of Sidaohezi Town from 1975 to 2015.This study provided a scientific basis for the selection of time interval in land use change simulation. Moreover, the simulation accuracy depends not merely on the land use transition probability within a certain period of time, but the local land policies and social development characteristics in the same time should also be considered.

Key words: Manas River Basin')">">Manas River Basin, land use and cover change, CA-Markov, time span