Research On Unified Simulation Method On Forest Optical And Microwave Signals Based On 3D Remote Sensing Mechanism Model RAPID2

As an important part of terrestrial ecosystems,forests are of great value to the carbon cycle and climate change of global terrestrial ecosystem.It is a common method to use multi-source remote sensing data to monitor forest vegetation growth,disturbance and restoration.Among them,the combination of optical and microwave remote sensing data can overcome the influence of cloudy and rainy weather,enrich the monitoring content from the time and space dimensions,and improve the accuracy and speed of large-scale forest vegetation parameter estimation.However,the inherent fusion mechanism of optics and microwaves data is still unclear,which limits the large-scale promotion and precision improvement of fusion.It is very necessary to use the remote sensing mechanism model to simulate the parameters and structure.RAPID2,a 3D remote sensing mechanism model,is developed based on computer graphics to construct 3D scenes to simulate the optical radiative transfer processes.RAPID2 is the best model for multi-source remote sensing data simulation based on a unified 3D scene definition.However,whether the input parameters of the model are consistent will have a fundamental impact on the simulation results.According to the allometric growth relationships of different forest types or tree species,the stand structure parameters corresponding to other variables of the forest's different age levels,different tree heights,etc.can be obtained.The inherent correlation of multiple variables can be established,which will effectively reduce the uncertainty and inconsistency of parameter input.This study focused on how to drive RAPID2 based on the allometric growth equations to conduct a unified simulation of forest optical microwave signals.First of all,three typical forest types of allometric growth equations were constructed.Secondly,the forest parameters of the same optical and microwave band were calculated by allometric growth equations,and the multi-source remote sensing signal simulations were carried out by running RAPID2 model.Thirdly,sensitivity analysis and multiple stepwise regression analysis were carried out for biomass parameters and multi-source remote sensing signals.Based on those analysis,the optimal allometric growth equations were filtered,which produced simulation results most in line with the actual situation.Finally,the accuracy assesment was conducted by comparing RAPID2 model simulations to Landsat 8 OLI,Sentinel-1B remote sensing data and the plot data of Gen He fired patch.Based on the above research contents,this paper realizes the simulation of different forest stands and the unified 3D structure scene of the visible light band,near infrared band spectral reflectance and C band,L band and P band backscatter coefficient simulation;finally,the allometric growth equation closest to the real forest information is screened out,and the method of the 3D remote sensing mechanism model to simulate different forest types is obtained.The results show that: the three groups of allometric growth equations of coniferous forests can be well applied to the study of the variation of canopy reflectance with LAI;the three groups of allometric growth equations of broadleaf forests are only suitable to simulate LAI and corresponding canopy reflectance values under large observation angles;the three groups equations of broadleaf forests can simulate reflectance varying with tree heights at different observation angles,but can not be sued for multi-source data fusion analysis;the three groups of allometric growth equations of mixed forests are suitable for studying the influence of observation angle on the relationship between spectral reflectance and LAI,however,not suitable for simulating the backscatter coefficients varying with LAI,thus not suitable for multi-source data fusion research.As for the model validation,the fitting accuracy of Landsat 8 OLI image and tree height plots is the best(R2 = 0.581,RMSE = 0.0164),and the blue band(483nm)has a good linear regression to tree height;the Sentinel-1B data are linearly related to tree height,and the VV polarization has the largest contribution to tree height in different polarization modes at C band.In the fusion analysis of optical and radar data,the results of remote sensing and simulation show that the visible bands have a significant impact on LAI;while during analyzing the impact of fusion data on tree height,the contribution of Sentinel-1B in remote sensing data to tree height is greater than that of Landsat8 OLI.