Estimation Of Forestry Above-ground Biomass With Multisource Remote Sensing Data

Forest aboveground biomass is an important indicator for studying global climate change,sustainable forest management and ecosystem function,and is associated with many important components of the carbon cycle,soil nutrient allocation,fuel accumulation and habitat environment in terrestrial ecosystems.In recent years,with the development of technology,remote sensing has been more and more widely used to estimate forest AGB.However,there is a large uncertainty in the AGB results estimated by remote sensing methods.This paper selects the Daxinganling forest area of Inner Mongolia as a research area to conduct forest AGB estimation based on optical and LiDAR remote sensing data.The four-scale geometric optical model can simulate the relationship between canopy structure parameters and canopy reflectivity.In the four-scale geometric optical model,the vegetation canopy reflectance is simulated by setting the range of input parameters,and a lookup table between the canopy structure parameters and the canopy reflectance is established.Using the TM remote sensing number to retrieve the canopy structure parameters,and then estimate the forest AGB,and compare it with the sampled results and the results obtained by DVI method In order to further improve the estimation accuracy of AGB,the tree height information acquired by LiDAR data is combined with the inversion of canopy structure parameters using TM remote sensing data to establish a binary equation for estimating forest AGB.Evaluating the effect of combining tree height with canopy structure parameter remote sensing data to improve AGB estimation accuracy,and the regional mapping of AGB was carried out The main contents and conclusions of this research were as follows:(1)According to the observation data of coniferous and broad-leaved forest plots,the canopy surface area(SA)was calculated by using the length of the semi-axle length(r)and the short semi-axis length of the canopy(b),and then the model of single-wood AGB was estimated The model was validated using independent sample data The R2 of the coniferous and broad-leaved forest single-wood AGB estimated by SA was 0.34 and 0.58,both of which reached a significant level of<0.01.(2)Estimating Forest AGB Based on TM Optical Data.Firstly,a four-scale geometric optical model is used to perform a multi-variable forward simulation(MFM)to establish a lookup table for estimating the canopy parameters required for SA.Based on the TM image data,the parameters required for SA calculation are calculated,and then the forest AGB is estimated;The forest AGB remote sensing estimation results were validated using observation data from 78 plots in the Root River,Alongshan and Moerdaoo areas,and compared with the model estimation results based on the Difference Vegetation Index(DVI)method.The feasibility of using the four-scale geometric optical model to invert the forest AGB was evaluated.The results show that for the three regions of Genhe,Alongshan and Moerdao,the consistency between the measured data and the estimation results based on the MFM method(R2=0.45,RMSE=20.8t hm-2,R2=0.63,RMSE=32.71tthm-2,R2=0.54,RMSE=27.58tthm-2)is significantly better than the estimation results based on the DVI method(R2=0.09,RMSE=27.69 thm-2,R2=25,RMSE=40.98thm-2,R2=0.16,RMSE=32.42thm-2).(3)Estimating Forest AGB Based on TM Optical and LiDAR Remote Sensing Data.In order to further improve the estimation accuracy of forest AGB,forest heights inferred from LiDAR data are combined with SA obtained from TM data to estimate forest AGB and compared with estimation accuracy using only TM optical remote sensing data or LiDAR data.The results show that the combination of TM optical data and LiDAR data can significantly improve the accuracy of AGB estimation R2 is 0.66,R2 is only 0.41 using TM optical data estimation results,and R2 is estimated to be 0.6 using only LiDAR data.This paper proves the feasibility of estimating the AGB by using the 4-scale geometric optical model to establish a look-up table and using TM remote sensing data to invert the canopy parameters.The combination of TM optical remote sensing data and LiDAR remote sensing data to improve the accuracy of AGB estimation is evaluated,which has an important reference role in the development of forest biomass remote sensing estimation methods.