Leaf Area Index Inversion Of Grass In Qinghai Lake Basin Based PROSAIL Model

Grass is an important component of terrestrial ecosystems，which is a naturalbarrier to the Earth’s surface, providing raw material for livestock and habitat forhuman survival. Leaf area index is an indicator of crop conditions biophysicalparameters, and vegetation biophysical processes are closely linked. Therefore, it isan important significance to accurate access to grass leaf area index of forage yieldestimates Qinghai Lake basin.Paper selected for Qinghai Lake basin as the study area, using MODIS andLandsat-8remote sensing images, combined with field data collected leaf area index,the measured spectral data using PROSAIL model radiative transfer model based ongrass leaf area indices from remote sensing data. The thesis includes the followingaspects:1. the data acquisition and pre-processing: the measured spectral data, includingplots, leaf area index data, chlorophyll concentration data collection and collation;atmospheric correction Landsat-8remote sensing images.2. PROSAIL model simulation analysis: combining the measured spectralreflectance data into leaf canopy reflectance, and the applicability of the model grassPROSAIL were analyzed;3. PROSAIL model parameter sensitivity analysis: based on the measured dataanalytical sensitivity PROSAIL model input parameters. Sensitivity is calculatedbased on the model and determine the sensitivity of a given model parameters.4. looking to build tables: the sensitive parameters in accordance with a certainstep size values obtained canopy reflectance leaves under different circumstances, theestablishment of LAI and canopy reflectance lookup table;5.the leaf area index (LAI) inversion: will be too remote sensing of atmosphericcorrected reflectance pixel according to the cost function with the lookup table tomatch the look, the corresponding canopy leaf area index LAI, and then plot themeasured data with the anti-speech to verify the results. Through research, thefollowing conclusions: 1. PROSPECT inversion model in terms of the grass leaf area index has betterapplicability: grass leaf blade and the measured reflectance reflectance PROSPECTmodel inversion of the match better, both for the absolute deviation.2. PROSAIL model input parameters to low sensitivity to LAI> Cab> Cm> SL> N>Cw, LAI and Cab determine the two most sensitive parameter for establishing LAI-grass canopy reflectance lookup table, the corresponding select Landsat-8,bands4,5,6participation LAI image inversion，and select MODIS bands1、2.3.The inversion results with measured LAI has a good consistency,Ladnsat-8correlation coefficient are R2=0.855, RMSE=0.63；MODIS correlation coefficientare R2=0.809, RMSE=0.86；...