Modeling MODIS LAI Time Series Using Stochastic Approach

Leaf area index is one of the most important variables characterizing energy flux, atmosphere- land surface exchange and vegetation structure. In this research, Leaf area index time series of wheat, alfalfa, apple and potato have been investigated using the modeling with basis on Box-Jenkins method and with the result of mathematical relationships between leaf area index and NDVI were compared. The trend of MODIS time series in the period of 2012 to 2015 with seasonal length 46 using seasonal Man –Kendall test was negative. The simulation of time series using ARIMA (p,0,q)×(P,1,Q)46indicated the error decreasing and correlation and NASH coefficient for example the correlation coefficient of wheat, alfalfa, apple and potato was 0.9, 0.94,0.93 and 0.89 respectively. The criteria comparison demonstrated the better performance of SARIMA model especially for alfalfa and apple time series with high values of leaf area index. Stochastic modeling has minimum error relative to the mathematical relationships. According to the high performance of SARIMA model, the preparing of leaf area index time series with high precise has more importance.