Accuracy Comparison Of The Leaf Area Index Inversion Method Of Winter Wheat At Different Growth Stages Based On Landsat_TM Data

Leaf area index as an important parameter of the vegetation canopy, Accurately obtainthe leaf area index It is significance for gains in detection yield assessment, disastermonitoring of crop. the development of remote sensing technology and appear of remotesensing model provides a reliable way to solve problem. However, in many remote sensing ofleaf area index to determine the optimal inversion method has positive significance.Based on Landsat_TM data of Tongzhou and Shunyi district,in jointing stage,floweringstage and mature of winter wheat, using the vegetation index optimization, PCA, and aphysical model inversion the three fangfa to inversion the leaf area index of the threestage.Some improvement in the calculation on the basis of the three methods, Vegetationindex method uses the preferred method of index points from three aspects of the correlationbetween anti-saturation and index of its own characteristics to choose the best index toinversion tne LAI,The PROSAIL model inversion method based on certain trends in the errorbetween the actual value of the model system error correction to the inversion accuracy of thethree methods were analyzed on the basis of the above. In particular, the thesis also try basedon Matlab neural network toolbox of BP neural network method of leaf area index inversion,the results show that, the BP neural network model inversion results and the actual leaf areaindex data the change tendency is consistent.The inversion results show that traditional empirical statistical methods as a simple andwidely used in the leaf area index inversion method is still greater advantage, In this article,the traditional vegetation indices and principal component analysis in the jointing stage andthe flower stage respectively obtain the best results, and the MSAVI and the second principalcomponent (PCA-2) TM in the two periods of inversion accuracy of82%and80%.Atmaturity, due to the water portion of the chlorophyll loss and changes in chlorophyll content, resulting in canopy spectral changes, about60%-70%in the retrieval accuracy of the periodof experience in statistical methods, it is not gond enough，during this period method thatfollow certain physical relations and the close physical contact with the leaf biochemicalparameters model (PROSAIL model) method has certain advantages. After joining the leafarea index obtained by the inversion error correction,The inversion result accuracy of76%-80%.Achieved a higher accuracy than traditional methods.Therefore ultimately determine theoptimal inversion method for the three periods were: vegetation exponential inversion(MSAVI), principal component based on Landsat_TM data analysis methods (principalcomponent), and after the error correction PROSAIL physical model inversion method, theinversion method to determine the overall increase in the efficiency of the inversion of leafarea index.