Estimation Of Vegetation Fractional Coverage Based On ZY-3Satellite Data

Vegetation fractional coverage is a comprehensive quantitative indicator tocharacterize the vegetation coverage. It’s not only an important part of the eco-climate model and eco-hydrological model, but also an important role of dynamicmonitoring in soil erosion and land degradation. Research on vegetation fractionalcoverage has important implications for the global environment change, globalwarming and other issues.The research, quantitative estimate of vegetation fractional coverage in the studyarea, aims to analyze and explore the feasibility and replicability of ZY-3on aquantitative estimation. During the study, in order to ensure the estimation accuracy,we investigate the data pre-processing of ZY-3in detail. Including radiometriccalibration,calculation of apparent reflectivity, atmospheric correction of ZY-3. Inaddition,different atmospheric correction methods were compared and analyzed, andfinally get the most suitable atmospheric correction method of ZY-3data. Based onthe dimidiate pixel model, which is the most widely used, most studied,and mostaccurate method, we select the effective remote sensing of NDVI. Visual estimationmethod is combined with the camera to verified the estimation accuracy. The mainlyresearch contents and methods are as follows:(1) Apparent reflectivity of ZY-3. As an apparent reflectance calculation parameters,the impact of ZY-3’s EBMSIis crucial. This article responds by WRC solar spectralcurves and ZY-3spectral function curve, using the integral equation, completedthe calculation of the parameters.(2) Atmospheric correction methods of ZY-3. For quantitative remote sensing, theimpact on the accuracy of surface reflectance estimates that can not be ignored. Inthis paper, FLAASH,6S, atmospheric adjacency effect correction,three methodswere used for ZY-3’s atmospheric correction. Finally,we compared thecalibration results with the measured spectral data, then chose the best suitablemethod for atmospheric correction.(3) Building of dimidiate pixel model. The level of estimation accuracy largelydepends on the choice of estimation model. We chose dimidiate pixel model forthe estimation of vegetation fractional coverage, which is high estimationaccuracy, wide range of applications. For the parameters of the model, first, wedefined a confidence size, then we calculated the paraments of full vegetationcover NDVIveg and naked soil cover NDVIsoil one by one,the basis were landuse type, and soil type, finally we got the estimation results of vegetation fractional coverage.(4) Estimation accuracy verification. The standard to evaluate the vegetationfractional coverage estimation accuracy is to verify the accuracy of estimationresults. In this paper, the measured data of visual estimation and photographicmethod were used to verify the estimation results, finally affirmed the estimationaccuracy of ZY-3data.(5) The feasibility of quantitative remote sensing on ZY-3. As China’s first domesticsatellite mapping data, use and promotion ZY-3data is not perfect, a lot ofresearch methods and applications are still under study. Based on the contrast ofthree atmospheric correction data and the estimation results of vegetationfractional coverage, we verify the reliability of ZY-3data, as well as thefeasibility of quantitative remote sensing. What’s more, this paper can providepractical support for the promotion and use of ZY-3data.