Retrieval Of FPAR Based On Energy Conservation Principle Using Remote Sensing

The fraction of absorbed photosynthetically active radiation (FPAR) is a physiologicalparameter describing the vegetation structure and the exchange of carbon and energy betweenvegetation canopy and atmosphere. It has been recognized as one of the essential climatevariables for global climate change studies by the Global Terrestrial Observing System(GTOS) and the Global Climate Observing System (GCOS). It is not only considered as oneimportant input in light use efficiency model and crop yield estimation model, it is also playsa key role in vegetation phenology detecting and crop condition monitoring. Thus, it isinteresting for estimating FPAR using remote sensing and has important implications forglobal change research.In this paper, a simplified semi-empirical model of estimate FPAR base on principle ofenergy balance and non-linear mixing pixel model (FPEB) was presented firstly. It has beentested for analyzes its feasibility and reliability using the ground data of different independentexperimental area. And the FPEB model also was compared with FPAR-NDVI statistic modelthen they are use to remote sensing. Through choosing the uncertainty factors, the sensitivityanalysis is used to determine the influence of input parameters and variables to FPEB model.Specific research includes the following three aspects:1. A method for retrieving vegetation canopy reflected and soil absorbed radiationcomponents of incident PAR was analyzed using non-linear mixing pixel model, and asimplified semi-empirical model of estimate FPAR base on principle of energy balance (FPEB)was presented firstly. Then, the FPEB model were examined using four independent field atNagqu district, Tibet Autonomous Region in2011and2013, Dangxiong county, TibetAutonomous Region in2011and Hailar district, Inner Mongolia Autonomous Region in2013. And the FPEB model was also compared with to the traditional normalized differencevegetation index (NDVI) model. The results showed that the accuracy of FPEB model wasbetter than NDVI-based model, and that the FPEB model was universal and valid for differentregions or vegetation types.2. Conduct avation remote sensing image FPAR inversion of airborne remote sensingexperiment in Qiangtang Plateau and heihe watershed allied telemetry experimental research（HIWATER）, and the result was compared with the result of FPAR-NDVI model inversionFPAR, it was construted using the ground data of airborne remote sensing experiment inQiangtang Plateau. The results show that the results of FPEB model inversiong FPAR andFPAR-NDVI model retrieving FPAR has perfacter consistency, but inversion accuracy ofFPEB model is better than the inversion accuracy of FPAR-NDVI model.3. Through choosing three uncertainty factors (solar zenith angle, vegetation canopytransmittance, vegetation coverage), the sensitivity analysis is used to determine the influenceof input parameters and variables to FPEB model. Results showed that vegetation coverage isthe important influence factot. Followed by solar zenith angle, with it increases, its sensitivityis reduces to FPEB model. Then sensitivity is beconing saturation when the values ofvegetation coverage are lager than80%. The sensitivity of vegetation anopy transmittance islow than solar zenith angle to FPEB model, its variation tendency is consistent with solarzenith angle.