Evi

Summary: The North American Monsoon (NAM) dominates summer climate and is responsible for providing the majority of rainfall over a large portion western North America. Knowing how ecosystems respond to this input of rainfall is critical to understanding how surface vegetation response may affect monsoon intensity. In this study, we use satellite observations to quantify the surface temperature and vegetation response of ecosystems to the monsoon rains along a latitude gradient that varies in monsoon intensity. The onset of the NAM rainy season in early summer is one of the main precursors for vegetation growth across the entire NAM region. At all the sites, it is observed that the mean daytime land surface temperature cools several degrees as the NAM fully develops. We also found that desert shrub and subtropical shrub regions in northern latitudes show large temperature and smaller vegetation variations, while the tropical deciduous forests in lower latitudes show much larger vegetation and smaller temperature changes. Satellite data proves to be a valuable tool for assessing the dynamics of seasonal and interannual surface characteristics helpful in determining climate responses.

Climate variables and their fluctuations dramatically affect terrestrial ecosystems and their variations. Vegetation indices have been used in numerous studies to investigate the relationship between ecosystem changes and climate parameters. In this study, GIS based spatiotemporal analyses were applied to model the relationship between vegetation variations based on the EVI-MODIS and its response to land surface temperature (LST) and rainfall in Mazandaran province during the period of 2000-2016. The LST parameter was derived from the MODIS data and rainfall parameter was achieved via meteorological station data in the region. Correlation and linear regression analyses were used to study the relationship between spatiotemporal enhanced vegetation index (EVI) and two climatic parameters. The findings indicated that the EVI had a rising trend over the study period which was mostly due to the increase in paddy fields. There was also a significant spatial correlation between EVI and LST which was significant and direct in the winter months and reversed during summer. The tabulate area analysis showed that throughout the winter months the spatial distribution of the highest EVI pixels matched to the maximum temperature pixels (20 to 27 ° C), while during June to September, the maximum EVI values were related to the areas in which the LST was less than 25 °C. Although there was no significant relationship between EVI/MODIS and rainfall in studied area, they reached a peak with a lag time of 1/5 to 2/5 months in the spring. The final results showed that the temperature is the main EVI climate factor in region and MODIS products have high potential to reveal the spatiotemporal dynamics of vegetation, the impact of human factors and its relation with the climatic factors of temperature and rainfall in the region.
Extended Abstract
1-Introduction
Vegetation can be considered as a comprehensive index reflecting changes in its surroundings due to interactions with the climate of each region. Climate, as a controlling and dominant factor, not only plays a great role in controlling and distributing vegetation and its spatiotemporal variations, but also has interactive effects, which has been the subject of a majority of research in the world. Besides, the study of vegetation dynamics monitoring in recent decades has been widely considered based on the use of vegetation indices extracted from satellite imagery. One of the most important of these indices is EVI, which has a high sensitivity to structural variables of vegetation. On the basis, data from the MODIS satellite data and products by making time series of EVI were applied in the current study to determine the dynamics of this index relative to two main climate variables (land surface temperature and rainfall) over a period of time.
2-Materials and Methods
In this study, spatiotemporal analysis of GIS was used to model the relationship between vegetation dynamics based on EVI from MODIS satellite. In fact, we tries to investigate the response of this index to two climate variables, including land surface temperature (LST) -as an auxiliary characteristic for temperature- and rainfall in Mazandaran province in northern Iran during the period 2000 to 2016. The LST was derived from the 17-year data from the MODIS satellite products. The rainfall information for this province was extracted from the Kriging interpolation based on the meteorological stations data available in the region. The EVI time series was also obtained from the MODIS products for the study period. Mapping was prepared based on the 17-yearly average monthly EVI and LST parameters extracted from the MODIS satellite. Three classes of EVI values (EVI<0/2, 0/4>EVI class, 0/2<EVI<0/4) were used to simulate the EVI trend during the study period from linear regression analysis and to examine the spatial distribution pattern of this index over four time intervals (2000, 2005, 2010 and 2016). Pearson correlation analysis was applied between the 17-yearly average monthly climate variables and the EVI at level of 0.95 to show the correlation pattern of the spatiotemporal relationship between vegetation changes and these variables. Tabulate area analysis was used to analyze this relationship in GIS environment. OLS regression analysis was also employed to generate the scatter plot of the linear spatial relationship between EVI/MODIS and LST-MODIS and precipitation.
3-Results and Discussion
The EVI temporal pattern based on regression analysis showed that the overall trend of this index during the study period was linear which ascended with an average of 0.45. This trend has more fluctuations in the summer than in the spring. However, the spatial distribution pattern of the EVI based on the three vegetation classes showed that during 2010 and 2016, areas with 0/4 >EVI (forest zones) had a decreasing trend which is due to the fact that the area under cultivation of rice farms (0/4<EVI<0/2) has been increasing. Moreover, the findings indicated that there was a significant spatial correlation between the EVI vegetation index and LST values, which had a significant direct correlation in the winter months and a reverse relationship in summer months. Maximum LST values are observed from June to September. The tabulate area analysis showed that during the winter months, the spatial distribution of pixels with the highest EVI values with pixels with a maximum temperature (20-27 degrees) has been overlapped. However, in June-September, the maximum EVI values were related to areas in which the LST was less than 25°C. Although no significant evidence has been found in simulating the relationship between rainfall and EVI extracted from the MODIS satellite in the study area, there is a delay of 1/5 to 2/5 months between them in the spring until reaching the peak. Besides, in the cold season, maximum rainfall is observed in areas with lower temperatures, while maximum vegetation is observed in areas with higher temperatures.
4-Conclusion
The results of this study showed that vegetation dynamics in Mazandaran Province responded to the land surface temperature fluctuations and this response was positive in the cold season. However, this relationship has not been seen in the case of rainfall. The increase in EVI in spring and summer is due to the beginning of the growing season and warming of the region, the highest in the Hyrcanian forests. The results also show the effects of human intervention on the dynamics of the spatiotemporal pattern of the region's land covers which was based on an increase in the EVI area of rice fields and a decrease in forest area floor levels despite EVI increases during the study period. The final results revealed that the temperature is the main climate factor of the EVI in region and MODIS products have high potential to reveal the spatiotemporal dynamics of vegetation, the impact of human factors and its relation with the climatic factors of temperature and rainfall in the region.