Spatiotemporal variation of correlation between vegetation cover and precipitation in an arid mountain-oasis river basin in northwest China
Spatiotemporal variation of correlation between vegetation cover and precipitation in an arid mountain-oasis river basin in northwest China
复制标题
西北干旱山地绿洲河流域植被覆盖与降水相关性时空变化
DOI：
10.1016/j.jhydrol.2019.04.044
复制
发表时间：
2019-07-01
影响因子：
6.4
通讯作者：
Bao, Zhenxin
中科院分区：
文献类型：
--
作者：
Mo, Kangle;Chen, Qiuwen;Bao, Zhenxin
水资源对干旱地区的生态系统和人类社会的生存至关重要。降水对植被覆盖的影响，特别是在干旱和半干旱地区，受到越来越多的关注。尽管在干旱和半干旱地区(年降雨量分别为50至200毫米和200至400毫米)，降水与植被覆盖之间的相关性已有报道，但不清楚在极端干旱地区(年降雨量低于50毫米)是否也存在同样的相关性。以中国西北部典型的干旱山地绿洲水系为研究对象，利用降水量和归一化植被指数遥感资料，研究了不同时空尺度上降水量和归一化植被指数的相关性。结果表明，降水量和NDVI在空间上分布不均匀。1999年至2007年期间，降水量以每年1.1785毫米的速度下降，而在2008年至2015年期间，降水量以每年2.0516毫米的速度增加。除耕地面积略有增加外，大部分地区NDVI没有明显的时间变化趋势(趋势斜率=0.001，P<0.05)。整个研究区空间平均降水量与NDVI的时间序列之间的Pearson相关系数为0.46(P<0.01)，表明研究区降水对NDVI的影响较大。降水量与NDVI的相关系数在空间上随降水量的变化而变化。通过时空位相分析发现，下游水系的NDVI与上游地区当月和前两个月的累计降水量有显著的相关关系。研究表明，降水对植被覆盖的影响在空间上具有明显的阶段性差异，而在极端干旱地区，上游邻近地区的前期降水对下游植被覆盖的影响较大。因此，在分析干旱区植被动态时，应考虑降水影响的空间分布和前期累积。
Water resources are essential for survival of both the ecosystem and human society in arid regions. The impact of precipitation on vegetation cover, especially in arid and semiarid areas, has received increasing interest. Although correlation between precipitation and vegetation cover has been reported in arid and semi-arid areas (with annual precipitation from 50 to 200 mm and from 200 to 400 mm, respectively), it is unclear whether the same correlation holds in extremely arid regions (with annual precipitation less than 50 mm). This study used remote sensing data of precipitation and normalized difference vegetation index (NDVI) to investigate their correlation at different temporal and spatial scales in a typical arid mountain-oasis river system located in northwest China. The results showed that precipitation and NDVI are not evenly distributed in space. In the period 1999-2007, precipitation was declining at a rate of 1.1785 mm/year, while during 2008-2015, it has increased at a rate of 2.0516 mm/year. NDVI showed no significant temporal trend in most areas (trend slope = 0.001, Significance > 0.05), except for a slight increase in regions where cropland expanded. The Pearson correlation coefficient between the time series of spatial mean precipitation and NDVI over the whole study area was 0.46 (Significance < 0.01), suggesting strong effects of precipitation on NDVI in the study area. The correlation coefficient between precipitation and NDVI varied spatially with the precipitation pattern in space. Through analysis of phase difference in space and time, it was found that NDVI in the lower reaches of the river system was significantly correlated with accumulative precipitation of the current and previous two months in the upper reaches of the system. Our study shows that the impacts of precipitation on vegetation cover have a clear phase difference in space, and that in an extremely arid area, antecedent precipitation in an adjacent region upstream has strong influences on vegetation cover downstream. Thus, the spatial distribution and antecedent accumulation of precipitation impacts should be taken into consideration when analyzing vegetation dynamics in arid areas.