MODIS NDVI与MODIS EVI对气候因子响应差异

Differences between MODIS NDVI and MODIS EVI in Response to Climatic Factors

以喀斯特地区植被为研究载体,利用2001—2010年MODIS NDVI、MODIS EVI序列和气候资料,分析两种植被指数与同期及前1—9期水汽压、降水量、相对湿度、最高气温、最低气温、平均气温、露点温度、气压、风速、日照时数多个气候因子的相关性及响应差异,为合理选择植被指数序列对植被进行监测研究提供参考。结果表明:1喀斯特地区植被EVI和NDVI与多数气候因子显著相关,但二者对气候因子的响应有明显差异。除日照外,EVI与其他各气候因子相关性明显高于NDVI。与NDVI相关性较高的气候因子为日照、最高气温、平均气温,其对应的相关系数平均为0.616、0.535、0.474;而与EVI相关性较高的气候因子为水汽压、平均气温、露点温度,其对应的相关系数平均为0.857、0.844、0.839。2喀斯特地区植被EVI和NDVI对多数气候因子的响应存在明显的滞后性,较之于EVI,NDVI对气候因子的响应更为滞后。EVI与气候因子的相关系数最高时期多出现在前1期,NDVI与气候因子的相关系数最高时期多出现在前2期。两种植被指数对日照、风速的响应无滞后性。除相对湿度、日照、风速外,各站点两种植被指数与其他各气候因子在0—9期的响应趋势均较一致。

In this paper, taking vegetation of karst area as the research site, MODIS NDVI, MODIS EVI and climatic information of 2001 to 2010 are used to analyze the relationship between vegetation and climate changes. Vapor pressure, precipitation, relative humidity, maximum temperature, minimum temperature, mean temperature, dew point, air pressure, wind speed and sunshine hours are taken as climatic variables to explore the relationship between vegetation index and climatic factors over different stages. Then, differences between NDVI and EVI in response to climate factors of karst area are analyzed. The objective is to provide references for reasonable use of vegetation index to monitor vegetation changes. The results show that: 1) There is a significant correlation between vegetation index and most climatic factors for both EVI and NDVI in karst area, yet differences between NDVI and EVI responsing to climatic factors are obvious. The correlation between EVI and climatic factors are obviously higher than that of NDVI except for sunshine hours. The correlation coefficients between NDVI and climatic factors are much higher for sunshine hours, maximum temperature, and mean temperature, their average values of correlation coefficients are 0.616, 0.535 and 0.474 respectively. While the correlation coefficients between EVI and climatic factors are much higher for vapor pressure,mean temperature and dew point, their average values of correlation coefficients are 0.857,0.844 and 0.839 respectively. 2) There is a significant lag in the response of EVI to most climatic factors. Compared with EVI, lag is stronger for NDVI than that of EVI. The highest correlation coefficient between EVI and most climatic factors appear at the earlier stage of-1, while the highest correlation coefficient between EVI and most climatic factors appear at the earlier stage of-2. For sunshine hours and wind speed, no response lag of two vegetation indices exists. The response of two vegetation indices to climatic factors is consistent for all stations except for relative humidity, sunshine hours and wind speed.