三套NDVI长时间序列植被指数的对比——以玛曲为例  

Contrast of the Long-term NDVI /MODIS, NDVI /GIMMS and NDVI /NSMC: a Case of Maqu

作　　者：沙莎[1] 郭铌[1] 李耀辉[1] 韩涛[2] 

机构地区：[1]中国气象局兰州干旱气象研究所,甘肃省干旱气候变化与减灾重点实验室,中国气象局干旱气候变化与减灾重点实验室,甘肃兰州730020 [2]西北区域气候中心,甘肃兰州730020

出　　处：《干旱气象》2013年第4期657-665,共9页Journal of Arid Meteorology

基　　金：国家科技支撑计划(2009BAC53B02);公益性行业(气象)科研专项项目(GYHY201006023);科技部农业科技成果转化资金项目(2011GB24160005)共同资助

摘　　要：NDVI／MODIS、NDVI／GIMMS和NDVI／NSMC是时间长度不同、空间分辨率相差甚远的3套ND—VI数据集，如何集成应用这些不同时间长度、不同分辨率的数据进行相关研究，数据集间的比较是最基础的工作。本文以甘肃省甘南州玛曲县为例，用直方图、相关分析、趋势分析等方法研究了这3套NDVI产品数据集的相互关系。结果表明：1）NDVI／NSMC与NDVI/MODIS的直方图具有类似的图像分布特征，但是NDVI／MODIS数据分布范围更大；2）3套NDVI在数值上表现为NDVI／MODIS〉NDVI／GIMMS〉NDVI／NSMC；3）3套数据集空间图像特征一致，两两间均具有十分显著的空间相关性，其中1月份相对最弱，5、10月份最强，三者相比NDVIfNSMC与NDVI／MODIS的空间相关性更强；4）1—3月、5—8月及年均的NDVI／GIMMS与NDVI／NSMC值存在显著的时间相关性，但两者逐年变化趋势存在较大差别，两者气候倾向率相差最大的高达5倍之多。NDVIfNSMC数据集在处理过程中可能未进行大气订正及交叉定标，这是造成共同源的NDVIfGIMMS与NDVI／NSMC差异较大的重要原因。NDVI/MODIS, NDVI/GIMMS and NDVI/NSMC are three NDVI datasets, due to their different lengths of time series and the giant difference in spatial resolution, the fundamental work should be done about comparing these datasets before applying these da- m to carry out relevant research. This article took Maqu County of Gansu Province as an example, used the ways of histogram, correla- tion analysis, trend analysis and so on to research mutual relations of three NDVI datasets. The results are as follows ： 1 ） The histograms of NDVI/NSMC and NDVI/MODIS had the similar distribution feature, but the NDVI/MODIS data distributed more wider than that of the former; 2）The numerical value of the three datasets were： NDVI/MODIS 〉 NDVI/GIMMS 〉 NDVI/NSMC; 3）Three datasets had consistent spatial image feature, there were very significant spatial correlation between any two of them, and in January the correlation was relatively weakest while in May and October it was strongest, among these datasets, NDVI/NSMC and NDVI/MODIS had more stronger spatial correlation; 4）The numerical value of NDVI/GIMMS and NDVI/NSMC from January to March, May to August and an- num average values had remarkable temporal correlation, but both of them had big difference in yearly changed trend, and the differ- ence of the climate tendency rate of them reached up to 5 times. The dataset of NDVI/NSMC might miss the atmospheric correction and cross - calibration during the data processing, which was an important factor resulting in big difference for the homologous datasets of NDVI/GIMMS and NDVI/NSMC.

关 键 词：NDVI MODIS NDVI GIMMS NDVI NSMC 长序列 玛曲 

分 类 号：P467[天文地球—大气科学及气象学]