新疆植被动态格局及其对气候的时滞效应

为明确新疆不同植被类型对水热变化响应的滞后时间,对新疆植被活动及其与气候变化的响应进行分析,研究基于1982—2015年的新疆GIMMS NDVI（normalized difference vegetation index）数据集、CRU降水与气温数据集,采用Sen + Mann-Kendall趋势分析、时滞偏相关分析、GIS空间分析和数理统计等方法,给出了34 a新疆植被格局动态变化特征,以及植被NDVI与气候响应的关系,探讨不同植被类型对气候响应的时滞效应。结果表明：1）新疆地区植被分布呈现北疆高于南疆、西部高于东部的空间格局,34 a来,研究区植被整体上呈现“变绿”趋势,在环塔里木盆地绿洲和天山山脉北段NDVI显著增加,伊犁地区呈现退化趋势;2）在月尺度时间分辨率下,新疆有72%植被区域对降水的响应存在滞后性,平均滞后时间为1.1个月,有70%的植被区域对气温的响应存在滞后性,平均滞后时长1.4个月,植被与气候要素时滞偏相关系数越高的区域,响应速度越快,总体上看,新疆地区植被对降水更为敏感;3）不同类型植被与降水和气温的响应程度不同,在新疆地区降水是草甸、灌丛和针叶林的主要促进因子,气温对阔叶林的影响最强,不同植被与降水的时滞偏相关系数均高于气温,不同植被对气温的响应时间均长于降水。总体上看,新疆地区植被与降水的相关性更高,植被对降水的响应比气温更迅速。

Dynamic pattern of vegetation in Xinjiang and its time-lag effect on climate

Time-lag effect of vegetation refers to the growth and distribution patterns out of sync with the varying climate. Vegetation cannot adapt immediately the climate change, leading to adjustment time. Previous research has focused mainly on the overall growth of vegetation under climate monitor or changes in a single type of vegetation. It is still lacking in the time-lag effect of vegetation types and their response to the climate. Therefore, this study aims to determine the vegetation activity and its response to climate change. The lag time of different vegetation types was also clarified in response to the hydrothermal changes. The Normalized Difference Vegetation Index (NDVI) and meteorological grid data were monthly collected from 1982 to 2015. The GIMMS NDVI dataset was also selected, together with the CRU precipitation and air temperature dataset. Sen + Mann-Kendall trend analysis was then conducted to obtain the interannual change rate of vegetation NDVI in over the past 34 years. At the same time, a significance test was carried out on the NDVI change trend, in order to distinguish the significant improvement and degradation areas. The growth and degradation of different vegetation types were statistically determined after the test. The lagged partial correlation analysis was used to calculate the partial correlation coefficients between vegetation NDVI and precipitation and air temperature in the current month, the previous month, the first previous two months, and the first previous three months. Then, the partial correlation coefficients were also synthesized to obtain the maximum partial correlation coefficient. The lag time was set as the maximum partial correlation coefficient. The dynamic characteristics of the vegetation pattern change over the past 34 years were presented to determine the relationship between vegetation NDVI and climate response. The lag effect was obtained in the different vegetation types on climate response. The results showed that: 1) A spatial distribution pattern of vegetation was achieved in the higher vegetation in northern Xinjiang than that in the southern, and higher in the west than that in the east. There was an overall "greening" trend in the study area. Specifically, the significantly improved and stable unchanged areas accounted for 20.6%, and 65.7%, respectively. The areas insignificant improvement and degradation accounted for 3.3%, and 10.4%, respectively. There was no area with significant degradation. In the spatial distribution, a significant NDVI increase was found in the oasis around the Tarim Basin and the northern section of the Tianshan Mountains, whereas, the Ili region showed a trend of degradation. In vegetation types, the cultivated vegetation shared the largest improvement area, accounting for 75.3%, while, the degraded area of coniferous forest was the largest, accounting for 24.2%; 2) There was a lag response of 72% of vegetation areas to the precipitation at the monthly time scale, with an average lag time of 1.1 months, while the lag response of 70% of vegetation areas to air temperature, with an average lag time of 1.4 months. The higher the lagged correlation coefficient between vegetation and climate elements was, the higher the response speed was. Overall, the vegetation was much more sensitive to the precipitation. 3) Different types of vegetation varied greatly in response to the precipitation and air temperature. Precipitation was the main promoting factor for the grassland, shrubland, and coniferous forest, while air temperature shared the strongest impact on the broad-leaved forest. The lagged partial correlation coefficients between different vegetation types and precipitation were higher than those of air temperature. The response time of different vegetation types to air temperature was longer than that of precipitation. There were lagged effects in the response of different vegetation to both precipitation and air temperature. The correlation between vegetation and precipitation was higher in Xinjiang, indicating a more rapid response of vegetation to precipitation than that to air temperature.