Estimation of leaf area index in Zagros forests using Landsat 8 data

Leaf area index (LAI) is an important feature controlling many forest ecological processes like gaseous exchanges between the atmosphere and forests, photosynthesis, evapotranspiration, water and carbon cycles. The aim of this study was to evaluate Landsat-8 OLI data to estimate LAI of forests in Zagros forests in Marivan region. To determine the relationship between LAI and OLI data, 60 sample plots were defined with a dimension of 45 × 45 m. The LAI value within each ground sample plot was determined using hemispherical photography and Gap Light Analyzer software. Suitable enhancement techniques such as vegetation indices, Principle Component Analysis (PCA), Tasseled Cap transformation and fusion were performed. The spectral values of corresponding plots were also extracted. Based on correlation analysis and contrary to expectations, negative correlation was observed between LAI, which represents the amount of greenness, and reflections in the infrared band. This could be due to the accumulation of dust on the leaves. The relationship between the spectral values and ground LAI were analyzed using stepwise multiple regression with 75% of the samples. Validation of resulted statistical models using the rest of control plots (25%) showed that the linear models are more efficient. The simple linear model of LAI and SR vegetation index (R2=0.682, RMSE% =22.6% and Bias% =-1.83) and the multiple regression model of NDVI, SR and SAVI (R2=0.722, RMSE % =20% and Bias% =-0.04) were the best ones for estimating the LAI from satellite data. Further complementary studies for determining the amount of LAI underestimation by hemispherical photography, capabilities of other remote sensing data and analysis methods are suggested to enhance this approach to estimate LAI.