北京大学学报自然科学版 ›› 2017, Vol. 53 ›› Issue (3): 563-572.DOI: 10.13209/j.0479-8023.2016.124

上一篇    下一篇

2000—2014年呼伦贝尔草原植被覆盖度时空变化分析

彭飞1, 范闻捷1(), 徐希孺1, 刘星2   

  1. 1. 北京大学遥感与地理信息系统研究所, 北京 100871
    2. 安徽理工大学测绘学院, 淮南 232001
  • 收稿日期:2016-01-21 修回日期:2016-06-16 出版日期:2017-01-22 发布日期:2017-05-20
  • 基金资助:
    863计划(2012AA12A304)和国家重点基础研究发展计划(2013CB733402)资助

Analysis on Temporal-Spatial Change of Vegetation Coverage in Hulunbuir Steppe (2000‒2014)

Fei PENG1, Wenjie FAN1(), Xiru XU1, Xing LIU2   

  1. 1. Institute of Remote Sensing and Geographical Information Systems, Peking University, Beijing 100871
    2. Anhui University of Science & Technology, Huainan 232001;
  • Received:2016-01-21 Revised:2016-06-16 Online:2017-01-22 Published:2017-05-20

摘要:

以呼伦贝尔草原核心区的新巴尔虎右旗、新巴尔虎左旗、陈巴尔虎旗和鄂温克族自治旗为主要研究区, 基于 MODIS NDVI 数据, 利用像元二分模型反演得到植被覆盖度, 并结合土地覆盖分类产品, 构建2000—2014年研究区植被覆盖度时间序列。通过时间序列分析, 从不同的时间和空间尺度分析草原植被覆盖度的变化规律; 同时引入覆盖度异常变化点检测算法, 并结合该地区同期气象数据, 进一步探讨研究区植被覆盖度变化与气象因子之间的内在驱动力关系。结果表明, 植被覆盖度在空间分布上主要表现为: 从东往西依次递减, 特别是研究区西南部, 覆盖度最低15 年来研究区植被年际变化总体上呈现前 10 年下降、后5年缓慢上升的趋势。对植被覆盖度的异常变化进行分析, 结果显示: 返青期和枯萎期覆盖度的剧烈变化与温度的相关性较大, 生长旺季内(7—8)月覆盖度的剧烈变化主要与降水量有关。

关键词: 呼伦贝尔草原, 像元二分模型, 植被覆盖度(FVC), 时间序列, 气象因子, 植被时序变化点检测算法, 相关性

Abstract:

As the core components of Hulunbuir Steppe, Xinbaerhuyouqi, Xinbaerhuzuoqi, Chenbaerhuqi, and Ewenkezuzizhiqi were selected as the key study area. Pixel decomposition model was introduced to retrieve the fraction of vegetation coverage (FVC) utilizing the MODIS NDVI data set, and the time series of vegetation coverage was reconstructed. Then, the temporal-spatial changes of FVC time series for study region over the 15-year period from 2000 to 2014 were analyzed. Especially, the algorithm of checking abnormal-point was also introduced, combining the same period of meteorological data of this area to investigate the interior driving forces between FVC and meteorological factors. The FVC distribution of 2014 showed that the higher level vegetation coverage mainly distributed in the east of study area; on the contrary, the lower level of that mainly distributed in the west of study area. Especially, the lowest vegetation coverage existed in the southwest of the study area. Further study of FVC changes showed that the vegetation coverage of whole study area decreased in the first 10 years, while that increased slowly in the latter 5 years. Additionally, the abnormal points which occurred in green-up and green-end periods had much more significant correlation with temperature; while the abnormal points which occurred from July to August strongly correlated with precipitation.

Key words: Hulunbuir Steppe, pixel decomposition model, vegetation coverage (FVC), time-series, climatic factors, changed points checking algorithm of vegetation time series, correlation