Forest Biomass Estimation In Fujian Province Based On Optical And SAR Remote Sensing

Biomass is the total weight of one or more biological organisms in a certain time and certain space.Biomass reflects the ecological community's ability in using natural potential,and is one of the important indicators to measure the community productivity.Using Landsat8 OLI images and 296 survey samples in Fujian Province,we found that the leaf biomass is highly correlated with the slope of near infrared and red band.Therefore,the slope of the near infrared and red band reflectance can be used as an effective indicator for describing and estimating the leaf biomass of forest.We setup the spectral slope model to indirectly estimate the aboveground biomass of Fujian Province.We also used the HV backscattering coefficient of SAR data and the traditional vegetation index regression model to estimate the aboveground biomass for comparison study.Finally,we analyzed and validated the accuracy of the three methods.The main conclusions are as follows:1)The spectral slopemodel is suitable for the aboveground biomassestimation in Fujian Province.We firstly estimate the leaf biomass based on the spectral slope model,and then according to the relationship between the leaf biomass and aboveground biomass,indirectly estimate aboveground biomass in Fujian Province.The estimated results agree well with the in situ biomass,showing the method is good and feasible.2)The comparison study of the spectral slope model and traditional vegetation index model for the aboveground biomassestimation.We used the two methods to estimate the aboveground biomass of Fujian Province,and then carried out the accuracy validation.The RMSE from the spectral slope model for the aboveground biomass estimation of coniferous forest,broad-leaved forest,and mixed forest are 49.8315 t/hm2(R2 = 0.5465),45.1820 t/hm2(R2 =0.4901),and 41.5131 t/hm2(R2 = 0.3879),respectively.The RMSE from thetraditional vegetation index model are 51.4462 t/hm2(R2 = 0.1510),45.9891 t/hm2(R2 = 0.1952),and 46.6258 t/hm2(R2 = 0.1546),respectively.The result shows that the estimated accuracy from the spectral slope model is much higher than the traditional one.3)The comparison study of the spectral slope model and SAR HV backscattering coefficient model for the aboveground biomassestimation.We used the two models based on different remote sensing data to estimate the aboveground biomass of Fujian Province,and then did the accuracy validation.The RMSE from the former model for the aboveground biomass estimation of coniferous forest,broad-leaved forest,and mixed forest are 49.8315 t/hm2(R2 = 0.5465),45.1820 t/hm2(R2 = 0.4901),and 41.5131 t/hm2(R2 = 0.3879),respectively.The RMSE from the latter model are 50.4858 t/hm2(R2 = 0.3446),46.1820 t/hm2(R2 = 0.3780),and 42.5131 t/hm2(R2 = 0.2375),respectively.The result shows that the estimated accuracy from the spectral slope model is higher than SAR HV backscattering coefficient one.In this study,we used pixel unmixing model to select the pure vegetation pixels,and setup leaf biomass regression model using(NIR-R)/(CWNIR-CWR)spectral slope.We firstly estimated the leaf biomass from the spectral slope model,and then indirectly estimated the total biomass of Fujian Province.This method eliminates the influences of the non-vegetation information on the biomass estimation,and avoids the issue that optical remote sensing can only record the vegetation canopy information,and increased the estimation accuracy of forest aboveground biomass in Fujian Province,providing a new method for forest biomass estimation from remote sensing.