近30年新疆植被生长异常值时空变化及驱动因子

Spatio-temporal patterns and driving factors of vegetation growth anomalies in Xinjiang over the last three decades

作为陆地生态系统的主体,植被的时空变化深刻地影响着景观格局和生态功能,深入理解植被动态及其对气候变化的响应,对于提高对生态过程的认识、加强生态管理具有重要意义。在一致性检验的基础上,利用中分辨率成像光谱仪(moderateresolution imaging Spectroradiometer,MODIS)的归一化植被指数(normalized Difference Vegetation Index,NDVI)数据集将新疆地区全球检测与模型研究组(Global Inventory Modeling and Mapping Studies,GIMMS)开发的NDVI数据集的时间序列拓展到2012年,探讨了生长季和各季节植被绿度、气候异常值的动态变化,分析了植被对气候变化的响应。研究结果显示,区域尺度和像元尺度GIMMS与MODIS NDVI之间的一致性较强。1982—2012年,研究区域生长季和各季节植被绿度呈显著增加趋势,但生长季存在明显阶段性:1998年前后分别呈显著增加和显著减少,夏季与秋季与生长季类似,而春季则不存在变化趋势的逆转。NDVI呈正异常值的面积比例与区域尺度NDVI的变化趋势一致;极端异常值、较大异常值多呈明显减少趋势,而一般异常值多呈增加趋势,NDVI的变化倾向于逐渐平稳。区域变暖趋势显著,降水量略有增加,潜在蒸散发显著提高,而湿润指数变化不明显。气温、潜在蒸散发主要在春季、秋季促进植被生长,而夏季降水量、湿润指数对植被生长的调节作用更为突出。

As one of the main components of terrestrial ecosystems,vegetation plays a key role in landscape structures and ecological services. Monitoring vegetation dynamics and their responses to climate change is beneficial to understanding ecological processes and designing adaptive management strategies. The GIMMS NDVI dataset from 1982 to 2006 is available for long-term NDVI trend analysis,and is widely used at global,regional,and local scales. Comparisons betweenthe GIMMS NDVI dataset and new products retrieved from new sensors should be conducted to link existing studies with future applications of other NDVI products in monitoring vegetation activity change. The MODIS NDVI dataset is referred to as the successor and improvement to the GIMMS NDVI time series. Based on a data consistency test,the time sequences of the GIMMS NDVI dataset were extended to 2012 with the MODIS NDVI dataset（ 2000—2012）. Then,we analyzed the trends of the NDVI anomalies and climatic factors（ temperature,precipitation,reference crop evapotranspiration,and humidity index） during 1982—2012 for growing season,spring,summer,and autumn. In addition,the response of vegetation to climate change was explored. The results showed that the GIMMS and MODIS NDVI data were very consistent at a regional scale,whereas at pixel scales,the spatial pattern of correlation between GIMMS and MODIS NDVI was significantly different. There was highly significant positive correlation between GIMMS and MODIS NDVI in northern Xinjiang,and poor correlation in southern Xinjiang. The areas with poor correlation were mainly distributed in deserts and alpine regions. In general,the two datasets are consistent and can be combined to expand the length of the NDVI time series. The vegetation greenness in growing season,spring,summer,and autumn increased significantly from 1982 to 2012 in Xinjiang. The NDVI in growing season increased significantly from 1982 to 1998,then decreased significantly from 1998 to 2012; this trend was also observed in summer and autumn seasons. Similar to the changes in greenness at regional scales,the percentages of land areas experiencing positive anomalies also increased significantly during 1982—2012. The area bearing extreme and large anomalies（ both positive and negative） of vegetation greenness generally decreased from 1982 to2012,and moderate anomalies（ both positive and negative） mostly increased from 1982 to 2012. The NDVI change tends to be gradually stable in study areas. The temperature and reference crop evapotranspiration increased significantly,and precipitation and humidity increased slightly over the past 31 years for all seasons. Vegetation growth in growing season was inhibited by both moisture and thermal conditions in Xinjiang,but the responses of vegetation to climate varied seasonally.The thermal was the primary climatic driver of vegetation changes in spring and autumn,and water resource affected plant growth in summer. The reduction of NDVI from 1998 to 2012 in the growing season,summer,and autumn was mainly due to drought stress,which was strengthened by warming in all seasons,and by reduced precipitation in growing season,spring,and summer. The long-term and consistent NDVI datasets offer a cheap,verifiable,and viable way to quickly detect change in vegetation,supporting managers in their effort to design and apply adaptive management strategies.