Study On Estimation Of LAI Models Based On TM

Taking Sunwu County of Heilongjiang Province as the study area, the statistic regression models are established between Leaf Area Index (LAI) measured in field and the different forms of Vegetation Index (â…¥) derived from Landsat TM imagery data.Based on the R and R2 values of the models, select the best model for LAI inversion. Also studied the internal relations between effective LAI and real LAI, and to compare which is batter between effective LAI-VIS and real LAI-VIS Regression Model. The study has led to the following findings and conclusions:1 Usually adopt LAI-2000s to measure of effective LAI existing an obvious error, This text knot of tradition measures LAI method and modern calculator technique together, and computes true LAI with the biomass model, builds up a linear regression model, examines accuracy attain to 92.4%.2 Build up the regression model of effective LAI-VIS, true LAI-VIS, discover the relativity of true LAI-VIS model being bigger than effective LAI-VIS model, beneficially explaining the accuracy of LAI number is very important for estimating the LAI.3 During various of LAI-VIS regression models, the exponential model is the best model, secondly is multivariate linear regression mode, but the multivariate linear regression model will lead the simple L4â… -â…¥relations to a complicated ones. Therefore, exponential model can more directly reflect the relation of LAI-VIS.4 During various of LAI-VIS regression models, the R-squared of NDVI-LAI is biggest, the relativity is the tallest, secondly is RVI, so NDVI and RVI suits more to estimate LAI.5 Carry on an accuracy examination to all models, finally, getting a best estimate LAI model is the exponential model of true LAI-NDVI, and make use of this model to create a LAI hierarchy scheme.