植物生态学报 ›› 2016, Vol. 40 ›› Issue (12): 1245-1256.DOI: 10.17521/cjpe.2015.0389

• 研究论文 • 上一篇    下一篇

黑河中游荒漠生态系统归一化植被指数对降水的响应

李芳1,2, 赵文智2,*   

  1. 1中国农业科学院草原研究所, 呼和浩特 010010
    2中国科学院西北生态环境资源研究院, 中国生态系统研究网络临泽内陆河流域研究站, 兰州 730000
  • 出版日期:2016-12-10 发布日期:2016-12-30
  • 通讯作者: 赵文智
  • 基金资助:
    国家重点基础研究发展计划(973计划) (2013CB429903)和国家杰出青年科学基金(41125002)

Changes in normalized difference vegetation index of deserts and dunes with precipitation in the middle Heihe River Basin

Fang LI1,2, Wen-Zhi ZHAO2,*   

  1. 1Research Institute of Grassland Science, Chinese Academy of Agricultural Sciences, Hohhot 010010, China
    and
    2Institute of Northwestern Eco-environment and Resource, Chinese Academy of Sciences, Linze Inland River Basin Research Station, Chinese Ecosystem Research Network, Lanzhou 730000, China
  • Online:2016-12-10 Published:2016-12-30
  • Contact: Wen-Zhi ZHAO

摘要:

降水是荒漠生态系统主要的水分来源, 是植被结构和功能变化的根本驱动力。该研究以黑河中游砾质荒漠(砾漠)和沙质荒漠(沙漠)为研究对象, 基于2000-2012年中分辨率成像光谱仪(MODIS)获取的归一化植被指数(NDVI)数据以及日降水数据, 运用多元线性回归法, 分析了砾漠和沙漠植被生长季(5-9月)、生长季早期(5-6月)和晚期(7-9月)累积NDVI (NDVIINT, INT表示某时间段的累积值)对冷季降水(Pc, 前一年9月至当年2月累积降水)、暖季降水(Pw, 当年3月至8月累积降水)、前一年生长季NDVIINT (NDVIINT-pys)以及干湿气候期(干旱期: 2001-2003; 湿润期: 2004-2007年)的响应。研究结果表明: (1)砾漠植被生长季NDVIINT年际变化的影响因素排序为NDVIINT-pys > Pc > Pw, 沙漠植被NDVIINT则为Pw > NDVIINT-pys; 砾漠生长季NDVIINT早期NDVIINT年际变化的影响因素排序为NDVIINT-pys > Pc, 晚期则为NDVIINT-pys > Pc = Pw; 而沙漠生长季早期NDVIINT年际变化的影响因素为NDVIINT-pys, 晚期是Pw。(2)在干湿气候期内, 降水量并非是影响荒漠NDVIINT变化的关键因子。干湿气候期交替时, 砾漠NDVIINT较沙漠增加明显; 湿润期内, 湿润期持续的长短是影响两种生境植被NDVIINT的关键因子, 以沙漠较为明显。黑河中游砾漠和沙漠植被生产力对冷暖季降水及干湿气候期响应具有明显的差异, 但总体显示出荒漠植被生产力对降水响应具有滞后性特征。以上结论可为揭示荒漠植被生产力对降水的响应机理提供参考。

关键词: 累积归一化植被指数(NDVIINT), 冷季降水, 暖季降水, 干湿气候期, 滞后性, 生态水文

Abstract: AimsPrecipitation is the major water source for desert ecosystems, with its temporal dynamics significantly driving the changes of ecosystem structure and function in desert regions. The objectives of this study are to evaluate the changes in normalized difference vegetation index (NDVI) with seasonal precipitation and different climate years in two cover types (desert and dune).MethodsBased on the daily rainfall dataset of 2000-2012 in the middle Heihe River Basin in Northwest China and the NDVI extracted from the moderate resolution imaging spectroradiometer (MODIS) images, we performed linear regression analysis to examine the correlation of NDVIINT (accumulated NDVI) with precipitation in two cover types (desert and dune). Two measures of the precipitation are Pc (cool-season precipitation from last September to late February) and Pw (warm-season precipitation in between March and August), while NDVI was characterized with NDVIINT-pys (previous-year during the different climate years (dry: 2001-2003, wet: 2004-2007)) and the different periods of a growing season (i.e. whole growing season from May through September, early growing season in between May and June, and late growing season in between July and September).Important findings We found that: (1) the determinants of growing season NDVIINT and their order were NDVIINT-pys > Pc > Pw for the deserts, while the order was Pw > NDVIINT-pys for the dunes. The determinants and their order of NDVIINT in early growing season were NDVIINT-pys > Pc for the desert, while they were NDVIINT-pys > Pc = Pw for the late growing season. However, for the dunes, NDVIINT of the early and late growing season appeared determined by NDVIINT-pys and Pw, respectively. (2) During the dry and wet periods, precipitation was not a significant factor influencing NDVIINT for the desert and dune. However, significant increases in NDVIINTwere observed at dune under wet condition. With the wet years continued, the length of the wet years become an important determinant of NDVIINT at both cove types, particularly at dune. In addition, it appeared that different changes in NDVI with precipitation existed between the two cover types, but with very similar effects of time-lag. These findings provide useful references for further understanding the mechanisms of NDVI changes with precipitation.

Key words: accumulated normalized difference vegetation index (NDVIINT), precipitation of cold season, precipitation of warm season, dry and wet climate period, lag time, ecohydrology