Research On Retrieval Algorithm Of FAPAR Based On The Layer Model

The Fraction of Absorbed Photosynthetically Active Radiation(FAPAR)is a canopy optical parameter related to the vegetation strcture,which describes the fraction of solar incident Photosynthetically Active Radiation(PAR)intercepted and absorbed by the vegetation canopy in the spectral range from 0.4 to 0.7um.As an elementary variable in characterizing the vegetation ecosystem,FAPAR is thus widely used in climate,hydrological,agricultural,and ecological models and has been identified by the Global Climate Observing System(GCOS)and Terrestrial Ecosystem Observation(GTOS)as one of the key parameters to reflect global climate change.Therefore,the research on the retrieval of FAPAR is of great importance for obtaining high-precision remoete sensing products and then serving global change research.Aiming at the main problems of current global FAPAR satellite products,a method based on the layer model was proposed to retreieve black-sky FAPAR,white-sky FAPAR and PAR absorbed by green leaves(FAPARgreen).Furthermore,the upscaled method of FAPAR from the instantaneous observation value to the continuous observation scale was developed.In a word,this paper not only realized the production of global white-sky FAPAR for the first time,but also expanded the definition of FAPAR products with the improvement of the accuracy of global FAPAR remote sensing products.The main contents and the achievements of the paper are as follows:(1)The radiative transfer process of the solar radiation was studied based on a canopy-soil dual-source layer model.At the same time,a non-linear spectral mixtue medel(NSM)was designed to retrieve the soil visible(VIS)albedo.Then,a method based on the energy balance residual(EBR)principle was presented to retrieve the black-sky and white-sky FAPAR.The PRO SAIL modeled datasets,MCD15A2H and GEOV1 FAPAR products were compared with the EBR FAPAR products.The validation results indicate that the NSM and EBR method can precisely retrieve the soil albedo,black-sky FAPAR and white-sky FAPAR.Direct validation against observations at VALERI sites demonstrated that the EBR black-sky FAPAR product was more accurate and unbiased(R2=0.917,RMSE=0.088,and bias=-2.8%)than the GEOV1(R2=0.868,RMSE=0.105,and bias=6.1%)and MCD15A2H(R2=0.901,RMSE=0.096,and bias=7.6%)products.(2)Based on the dual-source layer model,a leaf-wood-soil triple-source layer model(TriLay)was proposed to analyze the contribution of the non-photosynthetic components in forest vegetation canopy,and an algorithm for retrieving the photosynthetic active radiation(PAR)absorbed by green leaves(FAPARgreen)was proposed.The LargE-Scale remote sensing data and image Simulation framework(LESS)simulated data was used for validation.The results show that the TriLay FAPARgreen has a higher consistency with the LESS simulated FAPARgreen(R2=0.937,RMSE=0.064 under balck-sky condition,R2=0.997 and RMSE=0.025 under white-sky conditon).(3)For the limitations of the current instantaneous FAPAR remote sensing products,the variation characteristics of the instantaneous black sky FAPAR with the solar zenith angle(SZA)and canopy structure were studied in this paper,and an upscaled method based on the cosine function of SZA was proposed for achieving continuous FAPAR from the instantaneous FAPAR values.Both the PROSAIL simulated data and the field measurememts were used to validate the accuracy of the upscaled method.The results imply that the instantaneous black-sky FAPAR can be accurately upscaled to the daily averaged FAPAR,and the instantaneous black-sky FAPAR over 9:30AM was found to best represent the daily averaged FAPAR.