The Vegetation Change And Its Relationship With Meteorological Factors In Heihe River Basin During The Recent25Years

Heihe River Basin, located in the arid region of Northwest China, is well recognized for its diversity of natural landscapes and its fragility of ecological environment and is very sensitive to climatic fluctuations due to the effect of multiple landforms and local climate. Thus, it is of significant importance to reveal the history of vegetation variations in Heihe River Basin for the maintenance of ecological equilibrium, the establishment of ecological barrier as well as the development of western regions in China.Based on the GIMMS/NDVI dataset from1982to2006and temperature and precipitation observation data from12meteorological stations within Heihe River Basin, we focus on the variations of vegetation cover in the past25a temporally and spatially by ways of difference method, averaging method, correlation analysis together with polynomial curve fitting. The correlation analysis between NDVI and meteorological factors shows as follows:(1)Time variation trends of NDVI in Heihe River BasinAlthough the past25a witnessed an upward trend in vegetation cover, the tendency was quite slight. Employing the simple linear regression (y=0.004x-0.64, R2=0.223) to cope with the dataset, we find that values of NDVI went up by nearly6.2%in the25a and the minimum NDVI occurred in1985with a value of0.1485, while the maximum NDVI happened in2002with a value of0.1735. The annual average values of NDVI in the upper, middle and down streams was0.1843,0.1634and0.0919. The variation tendency of the upper stream kept in step with that of the middle stream, increasing by4.98%and6.7%, respectively, which was higher than the growth of the entire basin. The growth of NDVI in the downstream of Heihe River Basin was highest, and the cumulative increase was9.24%during this period. Using the multivariate curve fitting for the yearly average NDVI, we find this period can be roughly divided into two stages:a) NDVI maintained a significant rise over the period1982-1992; b) NDVI leveled off and kept relatively high. The variations of NDVI in the basin exhibited a seasonal rhythm apparently, e.g. the growth of vegetation starts in spring, reaches a peak in summer and declined gradually in autumn. NDVI of four seasons in the downstream was lower than that of the upper and middle stream. The NDVI in summer simply reaches above0.1. By comparing the variations of monthly NDVI over the three periods, we claim that May see the most dramatic growth, followed by June and October. Moreover, the growth of NDVI in spring in the upper stream was higher than that of other seasons remarkably, indicating that quick warming in spring resulting from global change has advanced the dates of the growing season.(2) Spatial evolutions of NDVI in Heihe River BasinBy analyzing the correlation coefficient between NDVI sequence and years, NDVI went up evidently (R>0.39, P<0.05) in231pixels, among which the pixels around Yeliugou, Tuole, Zhangye and Jinta could pass the significant test of P<0.01and P<0.001. The areas where the vegetation cover declined mainly concentrated in the middle and western part of Qilian Mountains. After classifications of NDVI during the two periods, we find that the number of those pixels representing high vegetation covers increased slightly, but the absolute number in the downstream was still quite low. Additionally, the number of the pixels standing for non-vegetation zones declined accordingly. There were no other significant changes anyway.(3) Response relationship between NDVI and meteorological factors in Heihe River BasinThe response relationships between annual average NDVI and temperature are higher than precipitation in the12meteorological stations, indicating that temperature plays a more important role in the variations of NDVI for the interannual variability. The response relationship between annual average NDVI and temperature in the middle stream is higher significantly than that in the upper and down stream. The most dramatic response to temperature happened in Yeliugou (0.692) and Zhangye (0.572), while the highest response coefficients were simply0.300in Minle and0.278in Shandan, not even passing the significant test (P<0.05). The meteorological factors had no effect on the variation of monthly NDVI in the downstream. The relationships between monthly NDVI and meteorological factors varied accompanied by different regions and different seasons. The results point out that spring temperature has the most significant effect to the oases area in the middle stream and precipitation plays a relatively important role in Minle and Shandan, which are far away from the oases. In autumn, at the end of the growing season, there is no relationship between the variation of NDVI and temperature and precipitation. Likewise, we find that lags between the variations of NDVI in summer and precipitation and temperature in spring are evident.