Study of Vegetational Phenology by Using Global Remote-Sensing Data and Climatic Data Set

本文论述了利用遥感影像资料探测到的全球植被的季节变化，并探讨了与气候有关的原因。本研究使用的数据是1986-1988年10分钟空间分辨率的NDVI（归一化植被指数）文件和美国NGDC编制的30分钟空间分辨率的逐月温度和降水数据集。考虑地形效应，即递减率和dem，将气候数据内插到10分钟分辨率。并在一定的约束下，将所有数据内插到半月数据中，以保持相同的年平均值或总价值。利用新建立的水平衡和积雪融化模型，以半月为间隔计算每个像元的供水量、实际蒸散量、储水量变化和最终径流量。我们认为，该模型中使用的假设是在自然现象的范围内，通过比较计算的流域径流和观察到的河流流量，仍有办法改进模型中的假设值。本项目新开发的系统将半月实际蒸散发和VI值的地图以动画图像的形式显示。动画地图不仅可以快速感知植被和气候参数的时间序列变化，还可以快速感知植被和气候参数的空间变化，并对两者的空间相关性进行统计分析。我们发现，植被的绿度变化（鲜绿和落叶）是同时发生的，或者在超过实际蒸散的阈值时发生半个月的延迟。结果表明，全球植被的绿度既不是由温度和降水这两个独立的气候参数控制，也不是由两者的简单组合控制。

This report deals with the seasonal change of global vegetation detected by remotely sensed image data and pursues the cause in relation to climate. The data used in this study are NDVI(normalized vegetatin index) files for 1986-1988 in 10 minute spatial resolution, and monthly Temperature and Precipitation data set in 30 minute spatial resolution prepared by NGDC, USA. Climatic data were interpolated to 10 minute resolution by considering the topographic effect, that is, lapse rates and DEMs. And all were interpolated to semimonthly data under constraint to keep the same yearly average or total value. Water supply, actual evapotranspiration, water storage change and resultant runnoff for each pixsel were calculated at semimonthly intervals by a newly developed model of water balance and snow accumulation-melting. The assumptions used in this model are, we believe, in the range of natural phenomena and there remains a way to improve assumptive values in the model through comparing calculated basin runoff and observed river discharge. Maps of semimonthly actual evapotranspiration and VI values were showed as animated images by a newly developed system in this project. Not only time-serial but also spatial change of vegetation and climatic parameters were quickly percepted in the animated maps and then spatial correlation of the two were analyzed statistically. We found that vegetational greenness change (fresh green and defoliation) happens simultaneously or with semimonthly delay to crossing a threshold value of actual evapotranspiration. We conclude that greenness of global vegetation is under control of the unique climatic parameter, actual evapotranspiration, neither two independent parameters temperature and precipitation nor a simple combination of the two.