Investigation the Effect of Extinction Models in Forest Height Estimation Using Two Different Physical Models Based on PolInSAR Data

This paper studies the effect of using two physical models in forest height estimation based on PolInSAR technique according to the different extinction models. We compare the results of random volume over ground (RVoG) and varied extinction random volume over ground (VERVoG) models in this paper. The first model which is constructed based on mean extinction coefficient has been inverted by three-stage method. In this paper we propose new method to invert the VERVoG which uses linear model for varied extinction over the forest height. To examine and compare the results we used real PolInSAR data of the European Space Agency (ESA), BioSAR 2007 campaign. The repeat pass experimental SAR(ESAR) L-band data, acquired over the Remningstorp test site in Southern Sweden. The experimental results emphasize that the extinction model greatly affects the accuracy of estimated forest height. It shows that the root mean square error (RMSE) of the RVoG and VERVoG models inversion are 6.18 and 2.08 m, respectively. In this regards, the varied extinction along the vertical direction improves the results against utilizing RVoG model which ignores the variation of the extinction. Also, the proposed method which can solve all parameters in the VERVoG model is the unique geometrical solution without the need for initial value, auxiliary data or iteration procedure.

Keywords:

Extinction coefficient , Extinction model , Random Volume over Ground , PolInSAR data , Varied Extinction Random Volume over Ground