摘要
引入雨水利用效率RUE计算人类活动对植被的影响。通过比较不同合成时间长度(30～360d)下RUE标准差,发现4—10月(210d)的RUE标准差(0.124)最大。将4—10月的RUE中值,约0.484～0.602DN每毫米降水作为无人为扰动的RUE。基于NDVI与降水的线性关系,将年降水量乘以无人为扰动的RUE作为NDVI预测值,得到1999—2004年NDVI残差(预测值和影像值之差)空间分布。方差分析表明,不同地区NDVI残差差异显著,南部几个旗县NDVI平均减小0.114,中部和东北部旗县NDVI平均增加0.055,西北部地区NDVI平均增加0.76。利用社会经济发展、土地利用、降水以及NDVI等资料分析了NDVI残差的空间特性,得出NDVI残差是由于人类活动引起的。
Rainfall use efficiency （RUE） is defined as integrated NDVi divided by annual precipitation data. The optimum time length representing annual plant growth is decided by comparing standard deviation of RUE under different composite time length which is from 30 to 360 days. It is found that the maximum RUE deviation occurred in time span from April to October which is also the growing season in local place. RUE median in each year, that is about 0. 484-0. 606 DN per millimeter precipitation, is selected to represent the RUE without human disturbance. Based on the linear relationship between precipitation and NDVI, the value of RUE median multiplying precipitation is used to predict the expected NDVI corresponding with local climate. Residue between predicted NDVI and smoothed NDVI data is used to indicate human influence. Va- riance analysis demonstrates that human impact on vegetation in different area has a significant difference. In the southern part of Horqin region in Inner Mongolia, human activities averagely decrease NDVI by 0, 114, whereas human activities exert positive influence on vegetation in the middle-north part and the northwest part, and human activities slightly increase NDVI by 0. 055 and 0. 076 respectively. Finally, economical development, land use, precipitation and NDVI are incorporated to evidence the rationality of the method.
出处
《中国沙漠》
CSCD
北大核心
2010年第4期896-902,共7页
Journal of Desert Research
基金
国家基础研究发展规划"973"项目(2009CB421308)资助
作者简介
彭飞（1984-），男，陕西咸阳人，硕士研究生，主要从事沙漠与沙漠化研究。Email：pengguy02@yahoo．com
