LAI Inversion Of A.Richii Based On HJ-1B Satellite Imagery Data And PROSAIL Model

Leaf area index (LAI) is a key parameter involved in a variety of ecosystem processes, corp growth and forest productivity models, because LAI directly influenced many biophysical processes, such as light interception and absorption, plant respiration, rain interception, carbon cycling, enery balance and atmosphere exchange between land surface and atmosphere boundary, LAI is one of the important variables indicate plant canopy structure. Therefore, LAI parameter acqusition has great meaning in terrestrial ecosystem research scope, but how to quickly get real-time LAI data has become research hotspot issue. Traditionally, LAI measurement method based on field observation only suitable for small area, but can’t satisfy regional and global LAI dynamic change research, remote sensing technology and radiative transfer theroy improvement provides possibility for large scale area LAI estimation.This paper utilized Look-Up Table method to inversion LAI based on HJ-1B CCD imagery data and PROSAIL rediative transfer model inversion. In order to get forest reflectance data from HJ-1B CCD images dense vegetation algorithm has been used, namely, using the spectral features relationship between1.66and0.66μm, estimate surface reflectance at0.66μm spectral band from top of atmosphere reflectance at1.66μm band firstly, and utilizing meteorological observation data, estimated surface reflectance at0.66μm band in dense vegetation area and different aerosol optical thickness variables excuting6S model in successive way to construct Look-up table. Through the apparent reflectance at0.66μm band and the cost function to minimizing the search process and obtain the aerosol optical thickness at0.55μm, then realizing atmospheric correction for each band, and get ground reflectance data in entire study area.In order to use PROSAIL model and look-up table techniques for LAI inversion, it is necessary to know the model input parameters on sensitivity of canopy reflectance. In this paper we had conducted sensitivity analysis via continuous change the parameters for the model. Sensitivity analysis result indicating that LAI, leaf mesh structure parameters, leaf slope angle influences the canopy reflectance in the whole solar spectral reflectance range, hot spot parameters, equivalent water thickness and dry matter content only significantly influence near infrared and outside band, conversely, the content of chlorophyll and carotenoids only affected visible band reflectance, and also analyzed canopy reflectance angular distribution characteristics at red and near-infrared wavelengths. According to the sensitivity analysis results and the measured spectral data, we determined the constant and free parameters of PROSAIL model and thus build the look-up table database. In LAI inversion cost function minimization method is used for search look-up table, the inversion results showing that the inversed LAI and the measured effective LAI from LAI-2000instrument is more identical, but the measured LAI is greater than the inversion results entirely.