New inversion approach for robust stem-volume retrieval from L-band SAR measurements with semi-empirical boreal forest model

This paper discusses stem volume retrieval based on inversion of SAR data, and proposes a new robust inversion approach. The method used employs model fitting with an inverted semi-empirical boreal forest SAR model, and takes advantage of the multitemporal aspect in order to improve the stability and accuracy of stem volume estimation. The approach is demonstrated using multitemporal dual-pol SAR imagery acquired by ALOS PALSAR sensor during the summer-autumn 2007. Reference data used for training the model included stand-wise forest inventory data from two boreal forest sites situated in central Finland. Multitemporal combination of model output in a multivariate regression framework allows volume estimates to be obtained with an RMSE about 43% of the mean of 110 m /ha, and a coefficient of determination R of 0.71. The methodology used can be employed to produce large-area stem volume maps from dual-pol ALOS PALSAR imagery mosaics.