Predicting forest changes based on climatic factors by using satellite images

Cover vegetation is much related to climatic conditions. Recognition of seasonal variability the vegetation growth is important to determine the ecosystems response to climate change in seasonal and interannual scales. In this research, to produce the predicted model, we used from climatic factors (precipitation, temperature and relative humidity (maximum, mean, minimum) during 20 years (1987-2006), and these data (141 weather stations) interpolated. So, calculate monthly the maximum value composite of NDVI from NOAA-AVHRR images that same during. Then, have computed Multivariate Least Squares regression from climatic factors (independent variables) and NDVI (dependent variable). The result show, most correlation between climatic factors and NDVI is about of 0.82 and occurring in May, that is highest growth. The lowest correlation is occurring in winters, because is not growing. The annual correlation in calculated model is more than 0.93 with inter the accidental errors. Totally, the calculated NDVI for May and Jun in during 2004-2005 years is close to predicted model, but in winters are distances from together. For the severe drought are lower predicted in 2006 at spring (Jun). In winter the role of temperature is more than precipitation and relative humidity in predicted model, but in earlier of May the role of precipitation and relative humidity is positive and temperature is negative, because is increasing temperature and decrease of precipitation and relative humidity. In autumn, is increasing the role of temperature and decrease the role of precipitation in predicted model.