中文摘要：

NDVI时间序列数据已经得到了广泛应用。中国的环境与灾害监测预报小卫星（HJ星）装载的4部CCD传感器能够满足在中国及其周边地区植被动态监测时间与空间分辨率两方面的需求,基于其建立的NDVI时间序列产品对于研究该区域的植被变化状况具有重要意义。然而波段的差异对不同传感器获取的NDVI具有非常显著的影响。本研究立足于对HJ星4部CCD传感器以及LandSat卫星TM、ETM＋传感器的波段分析,利用同时相遥感影像对获取并推算6部传感器NDVI在中国北方中温带干旱半干旱草地的NDVI交叉校正系数,并通过野外实测光谱数据计算等效NDVI对研究结果进行验证。结果表明,HJ星4部CCD传感器间获取的NDVI差异能够达到5%,本研究中得到的传感器间NDVI交叉校正系数能够将该差异有效减小到2%以内。同时证明了传感器中心波长对NDVI的影响大于波段宽度对NDVI的影响。

英文摘要：

The time series are being widely used.The Chinese HJ satellites which loaded 4 CCD sensors are capable of providing high spatial and temporal resolution observation data,which can establish the NDVI time series for environmental monitoring,disaster control and resources investigation,provide great opportunities for researches around China and its neighbouring areas.However,the band differences between sensors can significantly affect the NDVI accuracy.This paper based on band analysis of the 4 HJ CCD sensors and LandSat TM,ETM＋,using simultaneous image pairs,obtained the NDVI cross-calibration factors for the 6 sensors of mid-temperate arid and semi-arid grassland in northern China.By verification with equivalent NDVI from our field investigation in Hulunbuir,the result shows that the difference between sensors can reach up to 5%,which can effectively be reduced to 2% by applying the cross-calibration factor.Meanwhile,we proved the influence of NDVI from the center wavelength is much more than the band width.