草地学报 ›› 2021, Vol. 29 ›› Issue (S1): 199-207.DOI: 10.11733/j.issn.1007-0435.2021.Z1.023

• 技术研发 • 上一篇    下一篇

三江源地区土壤水分的地理分区遥感模型构建及时空变化

陈国茜1,2, 祝存兄1,2, 李素雲1,2, 周秉荣1,2, 李甫1,2, 曹晓云1,2, 周华坤3   

  1. 1. 青海省防灾减灾重点实验室, 青海 西宁 810001;
    2. 青海省气象科学研究所, 青海 西宁 810001;
    3. 中国科学院西北高原生物研究所, 青海省高寒区恢复生态学重点实验室, 青海 西宁 810008
  • 收稿日期:2021-03-01 修回日期:2021-05-14 出版日期:2021-10-30 发布日期:2021-11-17
  • 通讯作者: 李素雲,E-mail:lisuyun_qh@163.com
  • 作者简介:陈国茜(1986-),女,广西北海人,硕士研究生,高级工程师,主要从事RS与GIS技术在高寒生态气象中的应用研究,E-mail:guoxi_chen@163.com
  • 基金资助:
    青海省科技成果转化专项“青海重大气象灾害智能格点化防控技术提升与示范”(2018-SF-142);公益性行业(气象)科研专项“干旱气象科学研究——我国北方干旱致灾过程及机理”(GYHY (QX)201506001);2020年中国气象局兰州干旱气象研究所基本科研业务费项目“青海高寒草地夏季土壤水分可见光遥感监测技术优化及被动微波土壤水分技术研发”资助

Remote Sensing Model Constructions and Spatial-temporal Changes of Soil Moisture in the Three-River Headwaters Region

CHEN Guo-qian1,2, ZHU Cun-xiong1,2, LI Su-yun1,2, ZHOU Bing-rong1,2, LI Fu1,2, CAO Xiao-yun1,2, ZHOU Hua-kun3   

  1. 1. Key Laboratory of Disaster Prevention and Mitigation in Qinghai province, Xining, Qinghai Province 810001, China;
    2. Qinghai Institute of Meteorological Science, Xining, Qinghai Province 810001, China;
    3. Key Laboratory of Restoration Ecology of Cold Area in Qinghai Province, Northwest Institute of Plateau Biology, CAS, Xining, Qinghai Province 810008, China
  • Received:2021-03-01 Revised:2021-05-14 Online:2021-10-30 Published:2021-11-17

摘要: 为探求三江源地区土壤水分的时空变化规律,对研究区生态保护与修复、水源涵养能力建设、气候变化与灾害预警研究具有重要的指导作用与现实意义。本研究利用影响土壤水分变化的气候、植被和土壤因素对研究区进行地理分区,分析常用光学遥感干旱指数与土壤水分的相关性,基于中分辨率成像光谱仪(Moderate resolution imaging spectroradiometer,MODIS)数据构建土壤水分遥感监测模型,反演2003—2020年5—9月土壤水分,进行地面真实性检验,分析土壤水分时空变化特征及气象因子的影响。结果表明:地理分区的设置可减少大地形、植被差异对土壤水分遥感反演精度的影响,模型绝对误差为5.3%、均方根误差为6.8%;2003—2020年平均土壤水分为20%,以0.05%·a-1速率波动增加,且中北部和东北部局地明显趋湿、中部和东部局地变干;土壤水分变化受降水和气温的共同影响,尤其是降水影响明显。

关键词: 土壤水分, 地理分区, 遥感监测, 时空变化, 三江源

Abstract: To explore the spatio-temporal varation of soil moisture had an important guiding and practical significance for ecological restoration, water conservation, climate change and disaster warning in the Three-River Headwaters Region. In this study, the climate, vegetation and soil factors affecting soil moisture were used to take geographical division, and then the correlation between the drought index of commonly optical remote sensing and soil moisture was analyzed, on basis of MODIS data, a remote sensing monitoring model of soil moisture was consturcted to retrieve the soil moisture from May to September between 2003 and 2020, verify ground authenticity, and also analyze spatial-temporal changes of soil moisture and influences of meteorological factors. The results showed that, the geographical division could reduce the impact of large terrain and vegetation differences on the retrieval accuracy of soil moisture by remote sensing. The average absolute error and average root mean square error of the model was 5.3% and 6.8%, respectively. From 2003 to 2020, the average soil moisture was 20% with an increasing rate of 0.05%·a-1. In addition, the north-central and northeastern regions tended to wet, while the central and eastern parts became dry. Both precipitation and temperature affected soil moisture, in particular, the impact of precipitation was obvious.

Key words: Soil moisture, Geographical division, Remote sensing monitoring, Spatial-temporal changes, Three-River Headwaters Region

中图分类号: