Spatiotemporal Variation Characteristics And Impact Factors Of NDVI In Jiangsu Province

Vegetation is an important part of terrestrial ecosystem, it plays an important role in such aspects as climate regulation, maintaining biodiversity, soil and water conservation and water conservation. To recognize the vegetation variation characteristics and to explore the relationship between vegetation variation and climate change, can providing theoretical basis for dealing with the global change and they were of great significance in the study of global changes. Research in relationship between regional vegetation and climate factors can enrich the regional response of global change research, it also has reference value for recognize the regional ecological environment change trend and to formulate the reasonable ecological environment protection policies and measures.Normalized Different Vegetation Index reflects the vegetation growth status. Based on MODIS/NDVI data sets from 2001 to 2010 and meteorological data from 31 weather stations over Jiangsu province including temperature, precipitation, relative humidity and sunshine hours from 2002 to 2010, mean method, linear regression trend method, spatial interpolation method and correlation analysis method were used to analysis the spatial and temporal dynamic changes of NDVI from the overall regional and grid level, the relationship between NDVI and climate factors such as temperature, precipitation, relative humidity and sunshine hours were also analyzed in this paper. Main conclusions are as follows:(1) The monthly mean NDVI of Jiangsu province variation in the year present the double peak trends, the peaks were during April to May and during July to September. Inter-annual variability of NDVI in average between 2001 and 2010 is on the rise, and the trend of rising reached extremely significant level (P<0.01), the rising rate was 0.005 a year during the 10 years. Monthly, the rising trend of NDVI in May and October reached extremely significant level(P<0.01), the rising trend of NDVI in March reached significant level(P<0.05).(2) In Jiangsu province, the distribution and the change trend of average NDVI distinct in different districts. As to the distribution of average NDVI, the NDVI of some hilly areas were the highest, jianghuai plains and huanghuai plains took the second place, the NDVI of other areas were a bit low. As to the change trend of yearly NDVI, the vegetation of northern Jiangsu was improved and the vegetation of south-central Jiangsu remains unchanged. From the perspective of the inter-annual variability of monthly NDVI, the areas of NDVI change trend in May and October which reached significant level were 39.03% and 38.30%.(3) Temperature, precipitation, relative humidity and sunshine time all had significant effect on monthly change of NDVI, and the impact of temperature was the highest. Overall, the correlation between NDVI monthly change and temperature monthly change was the highest; precipitation, the second; and then the relative humidity and sunshine time. There exist lag effect and cumulative effect on response of NDVI to precipitation, the lag phase and the accumulation period both were 1 month, there exist cumulative effect on response of NDVI to relative humidity and sunshine time, and the accumulation period were two months, there were no lag effect and cumulative effect on response of NDVI to temperature.(4) The overall average annual NDVI in Jiangsu province on the inter-annual variability of climate response is not strong. The correlation between annual average NDVI and climate factors such as temperature, precipitation, relative humidity and sunshine time were 0.282,-0.142,-0.369,-0.004, all of them had not reached significant level(P>0.05)(5) In May, August and October, NDVI is strongly influenced by climate factors, the relationship between climate factors and NDVI of August is most closely. There were significant correlation between NDVI in August and temperature, precipitation, relative humidity and sunshine hours(P<0.05), and the NDVI in May and October had significant correlation with temperature and sunshine time(P<0.05).