气象因子与地表植被生长相关性分析

Correlation analysis between meteorological factors and the ratio of vegetation cover

降水和气温是影响地表植被覆盖状况的两个重要的气象要素 ,地表植被的生长对它们的响应又存在一定的滞后效应。利用我国 334个气象台站和 1 982～ 1 994年 NOAA- AVHRR卫星数据 ,通过相关分析研究了我国 4～ 7月份各气候区域降水和活动积温对地表植被的影响。结果显示 ,不同的气候区域和月份 ,降水和积温对地表植被的作用程度明显不同 ;前期降水和活动积温对植被作用的有效时间尺度也分布不均 ,并与土壤质地类型分布有关 ,因此 ,通过某一固定时间尺度的降水或活动积温不能准确的预测大区域地表植被的生长状况。

Rainfall and temperature are two important factors that influence vegetation coverage over the land surface. The up-growth of vegetation lags behind rainfall and temperature in general. Due to their inhomogeneous distribution, the effect of rainfall and temperature on land surface vegetation cover differs distinctly. Some research results show that the time lag is 14 days, and suggested that the NDVI (Normalized Difference Vegetation Index) could be used as an index for the reconstruction of rainfall in some regions. ;There are different climatic regions in China such as temperate zone, subtropical zone and tropical zone. The interactions between meteorological factors and vegetation in each climatic region are significantly different. Many factors such as the inhomogeneous distribution of rainfall and temperature, difference of vegetation types and soil textures play important roles in interactions. To develop advanced land surface feature models, these factors must be considered. ;Based on the daily mean rainfall and total active temperature from 334 weather stations and NOAA-AVHRR data from NASA during the period from 1982 to 1994, the influence of rainfall and total active temperature on vegetation in China is analyzed from April to July through correlation analysis. The Atmospheric correction scheme of channel one and channel two follows the algorithm of Gordon. NDVI is composited by taking the maximum NDVI value from the daily data within 10 days, with spatial resolution of 8km×8km.;The maximum NDVI within a 10 day period is assumed to equal the NDVI of the tenth day. This assumption also corresponds to the character of vegetation growth. First, rainfall or temperature is added retrospectively day by day, and then the correlation coefficients are calculated. The final outputs are maximum correlation coefficients and accumulative number of days while maximum correlation coefficient occurs. The result shows regional characters in each month. Vegetation does not grow in the north of China in April because of lower temperature. Correlation coefficient is low between NDVI and rainfall. Because winter wheat is growing then and rainfall is relatively lower. The regions with higher correlation are located basically in humid parts of the South Temperate Zone. In contrast, there is no evidence of higher correlation in the subtropics and tropics due to excessive rainfall and soil humidity although it is also being in the vegetation period. As temperature increases, the distribution of maximum correlation coefficients expends in May in northern China except for some regions with snow cover, where melting snow made soil much more humid; correlation in all other regions increases. At this time, the ripe winter wheat caused the correlation to decrease in the South Temperate Zone. Vegetation also does not depend on rainfall south of the Yangtze River because of overmuch rainfall. In June and July, vegetation demands much more rainfall in the middle part of the moist temperate zone, owing to up-growth. Cultivation of late autumn crops made vegetation more reliant on rainfall. In the middle and lower reaches of the Yangtze River, the advent of Meiyu results in no correlation between NDVI and rainfall. The temporal plot shows that NDVI lags behind rainfall with quite apparent differences in climatic regions where correlation coefficients are higher than 0.3. In drought regions, since a little rainfall could change vegetation cover significantly, the time lags are less than that in other regions. In rainy regions, precipitation supplies continuously water for vegetation growth and results in relatively long time lags. In April in the southern humid temperate zone with higher correlation, total rainfall during 45～65 days is best relative to NDVI. That means vegetation will be further affected if the total precipitation over 45～65 days changes. In May, in the mid-variable zones and semi-drought regions, vegetation is very sensitive to rainfall. As a consequence, rainfall within 15~45 days is most important for vegetation. In June a