Estimation Models of Forest Stand Biomass Using Combined Multi-Spectral and LiDAR Technologies
Investigating from Shimen National Forest Park,Conghua,Guangzhou City,Guangdong Province,four different forest types(mixed coniferous-broadleaved forest,broadleaved forest,coniferous forest,bamboo forest)were selected as study areas.Three 20 m×20 m plots were chosen for each forest type.Integrated with data from LiDAR,multispectral imagery,and field measurements,we developed multivariate nonlinear inversion models and multivariate linear regression models to estimate aboveground biomass.The optimal model was selected based on accuracy assessment.The results showed that:(1)The multivariate nonlinear inversion models based on multisource data achieved the highest accuracy for estimating aboveground biomass across the four forest types:mixed coniferous-broadleaved forest predicted biomass was 42.79 t·hm-2,broadleaved forest was 60.46 t·hm-2,coniferous forest was 32.99 t·hm-2,and bamboo forest was 1.92 t·hm-2.(2)Among the four forest types,the multivariate nonlinear inversion models showed decreasing fitting accuracy as follows:bamboo forest(R2=0.919),broadleaved forest(R2=0.813),coniferous forest(R2=0.786),and mixed coniferous-broa-dleaved forest(R2=0.713),all meeting accuracy requirements.Integrating multispectral and LiDAR data provided a more precise means to extract aboveground biomass information,facilitating accurate estimation across mixed coniferous-broad-leaved,broadleaved,coniferous,and bamboo forests.
NETL
NSTL
万方数据
