Forest Height Inversion Based On PolInSAR And TomoSAR

Forests are the largest terrestrial ecosystem on the planet and play an irreplaceable role in the carbon cycle and sustainable development.The spatial distribution information of forest height is of great significance for forest resource management,forest biomass estimation,and regional and global carbon cycle research.Forest ground survey data is limited in large-scale research applications.Microwave remote sensing is not affected by clouds and snow.It is capable of penetrating the forest at all times and all weathers.The backscattered signal contains information about the vertical structure of the forest and has other remote sensing incomparable advantage.Synthetic aperture radar(SAR)remote sensing has proven to be one of the most reliable techniques for accurate forest parameter estimation.Based on the simulation analysis of the forest scattering RVoG model and the relative error of the forest height,this paper used simulation data to perform a single-baseline PolInSAR forest height inversion,providing a basis for further forest height inversion using multi-baseline airborne PolInSAR data,using airborne TomoSAR data to inverse forest height profiles,laying the foundation for further forest heights inversion,large-scale forest aboveground biomass models and forest age estimation,and site quality evaluation.1.The RVoG model establishes a functional relationship between vegetation volume scattering coherence and vegetation parameters(forest height,extinction coefficient,ground-volume scattering ratio,topographic phase).The most important parameter is the forest height.This paper analyzed the effects of vertical wave number,2? hambiguity elevation,extinction coefficient,and ground-volume scattering ratio on vegetation volume scattering coherence,and simulated and analyzed the impact of time decoherence,vertical wave number,extinction coefficient,and ground-volume scattering ratio using a three-stage method on the relative error of retrieved forest height.2.Using the European Space Agency(ESA)Toolbox PolSARPro to simulate L-band(L=23cm)PolInSAR SAR data with an average vegetation height of 18m.This paper studied DEM difference,RVoG,SINC,Hybrid,and Hybrid method based on coherent optimization.Compared with the real value 18m,descending order of vegetation height means is SINC,Hybrid,improved Hybrid,RVoG,DEM difference method.The difference between the improved Hybrid inversion method and the real value is smallest of 0.12m,smaller than Hybrid of 0.31m.RMSE of the improved Hybrid,Hybrid,SINC,RVoG,DEM difference,is 1.06m,1.48m,3.49m,7.51m,and 8.04m.The vegetation height of the improved Hybrid method has the smallest difference and RMSE.Improved Hybrid inversion method produces the best result among the five methods,combining the merits of Hybrid with the coherent optimization,with the smallest difference between real value and RMSE of vegetation height and topographic phase.This paper improves the Hybrid method based on the coherent optimization and analyzes the accuracy of vegetation height with the ground phase estimation results to compare the five methods.3.A three-stage forest height inversion algorithm is studied systematically using ESA PolSARpro software to simulate the L-band(L=23 cm)full-polarization interferometric SAR data with an average forest height of 18 m.The three-stage algorithm is geometrical and the most widely used approach for forest height inversion based on the RVoG model using PolInS AR data.Comparing the three-stage algorithms without using coherence optimization methods(HV method)with the three-stage algorithms based on MCD coherence optimization methods.We introduce the coherence amplitude inversion method and modify the fixed weight to the variable of ? times the ground scattering ratio,improving the forest height inversion accuracy.The results show the average values of forest height from MCD and HV methods are 16.19 m,and 15.83 m,respectively.The RMSE from MCD and HV methods are 4.43 m and 4.80 m,respectively.For the methods introducing the coherence amplitude and modifying the constant weight into a variable,the average value of the inversion forest height from MCDWeight and HVWeight are 16.71 m,and 16.29 m,respectively.The RMSE of MCD Weight and HVWeight is 4.35 m and 4.65 m,respectively.It can be concluded that compared with the three-stage algorithm,a three-stage algorithm based on coherence optimization improves the accuracy of forest height inversion,introduces the coherence amplitude term,and modifies the fixed weight to the variable,further improving the precision of forest height inversion.4.To accurately inverse the forest height over a wider range,it is necessary to study the baseline selection method for multi-baseline PolInSAR data to alleviate large errors in forest height inversion over a wider range from single-baseline,exploring a better baseline selection method.The UAVSAR L-band multi-baseline full PolInSAR data was used from 5 orbits obtained by JPL/NASA in Pongara,Gabon forest on February 27,2016.Based on the maximum coherence difference(MCD)coherent optimization algorithm to make complex coherence maximum separation,the PROD method and the ECC method were improved,compared and analyzed;and verified using the LiDAR data LVIS RH100 obtained by NASA on March 4,2016.The error maps of the difference between the forest height and the LVIS RH100 inverted by the two baseline selection methods were plotted to analyze the results of the forest height inversion.And the density maps of the kz,canopy coherence amplitude corresponding to the two baseline selection methods and LVIS RH100 were plot to directly evaluate the difference between the ECC method and the PROD method selecting the baseline,and comparing and analyzing the pros and cons of the two baseline selection methods.Combined with these drawn graphs(forest height maps,error maps,density maps),the forest heights inverted by the two baseline selection methods were compared and analyzed.The error in low and high forest areas was large.The highforests were underestimated(the error was negative),and the forests in the low areas wereoverestimated(the error was positive).The underestimated or overestimation of the ECC method was greater than the PROD method,and the accuracy was inferior to the PROD method.The two methods had good consistency compared with the LVIS RH100 data.Validated with the LVIS RH100 data,the RMSEs of the ECC method and PROD method were 9.80 m and 8.86 m,respectively.The accuracy of the PROD method was improved by 9.63%than the ECC method.For low and high forest areas,when coherence separation or coherence amplitude between complex coherence was small,the degree of complex coherence separation and complex coherence amplitude were the main factor affecting forest height inversion.The PROD method comprehensive considered the complex coherence amplitude and coherence separation,so the PROD method was more suitable for inverting forest height than the ECC method.In medium-high forest areas,when the complex coherence was more dispersed and the coherence amplitude had not reached saturation,the straight line fitting effect of the coherence region was an important factor for forest height inversion.The ECC method was more suitable for inverting the forest height than the PROD method.Our results demonstrated that the forest height inversed by the two baseline selection methods was consistent with LVIS RH100.The ECC method only took the linearity of the coherent region as the criterion.Because of taking into account the fitting effect and coherence amplitude of the coherent line,the forest height inverted by the PROD method was better and closer to the LVIS RH100 than the ECC method,alleviating the phenomenon that the ECC method underestimates tall forests and overestimates low forests to some extent.The PROD method was more suitable for forest height inversion in low and high forest areas,and the ECC method was more suitable for forest height inversion in moderately high forest areas.5.Using the cropped data of BIOSAR 2008 activity to apply the Beforming and Capon methods in the TomoSAR technology to retrieve the forest height profile,and this paper compared and analyzed the two methods,and compared the information forest height profile in a single polarization channel with in the fully polarization channel.The results show that the classical Beforming method has low height-direction resolution and high side lobe levels,and the vertical structure is insufficiently estimated.The Capon method has ultra-high resolution and sidelobe suppression,but at the cost of loss of radiation resolution.The full polarization method considers the information of the three polarization channels comprehensively,and the obtained information is better than that of the single polarization channel.When the range direction is converted to the ground range direction and the DEM is removed,the forest height profile has better consistency with the forest height profile obtained by the LiDAR.