IWCM Model Based On Sentinel Data With Different Polarizations For Estimating Forest Above Ground Biomass

The forest ecosystem is the main body of the terrestrial ecosystem and maintains the global climate system.Forest plants absorb carbon dioxide and emit oxygen through photosynthesis,which makes forests play a very important role in the global terrestrial carbon cycle.Forest biomass is the basis for research on forest carbon sinks.Forest biomass refers to the total dry matter of the living organisms in the forest community within a specified period of time.It is equivalent to the forest coverage rate and is one of the important measurement indicators of forest resources.This study uses SAR synthetic aperture radar to estimate forest biomass.Compared with traditional optical remote sensing,SAR data has better penetrability and stability that is not affected by weather are its unique advantages.Accurate estimation of forest biomass can better understand the changes in forest ecosystems,and has a positive effect on forest management and scientific guidance of forestry management.Taking Maoershan Forest Farm in Northeast Forestry University as the research area,C-band sentinel dual-polarization data was used to estimate forest above ground biomass through the Interferometric Water Cloud Model(IWCM).Explore the relationship between coherence and backscattering and forest above ground biomass under different polarizations,respectively.And find a more suitable polarization mode for estimating forest above ground biomass.Simultaneous water cloud model and interfering water cloud model,using genetic algorithms to determine initial and non-linear least squares solution parameters.The relationship between backscatter,coherence,and forest above ground biomass was established for VV and VH polarization modes,respectively.Using the model established by backscatter,coherence and forest above ground biomass to invert forest biomass.Analysis of the backscattering coefficient and coherence coefficient of 43 plots in Maoershan showed that coherence is more sensitive to forest above ground biomass than backscattering,especially for the weakly penetrating C-band,there is no saturation phenomenon,and it is more suitable to use the interference coherence to retrieve the biomass.VV polarization estimation result R2=0.705,RMSE=29 t/ha;VH polarization estimation result R2=0.534,RMSE=40.48 t/ha.The VV polarization method has a higher R2 and a smaller RMSE,which is more suitable for inversion of forest above ground biomass.The VH polarization method is greatly affected by the attenuation of the vegetation layer,and the radar signal penetration is better under the VV polarization method,which is why the VV polarization method has a better inversion effect.The final results of this study show that the estimation of forest above ground biomass by interference coherence is a more stable way.Using the interference water cloud model based on Sentinel-1 data,forest above ground biomass can be well retrieved.The VV polarization method of Sentinel-1 data is more accurate and more suitable for forest above ground biomass estimation and regional forest above ground biomass mapping.