Pol-InSAR Approaches at L-Band: Actual Status and the Impact of Temporal Decorrelation

Abstract

Polarimetric Synthetic Aperture Radar Interferometry (Pol-InSAR) is a recently developed radar remote sensing technique, based on the coherent combination of SAR interferometry and radar polarimetry. The combination of polarimetric and interferometric information provides in unique way sensitivity to the vertical distribution of different scattering processes and makes the investigation of the 3-D structure of volume scatterers - by means of Pol-InSAR - a challenge. Indeed, in the last years, quantitative model based estimation of forest parameters - based on single frequency, fully polarimetric, single or multi-baseline interferometric configurations - has been demonstrated; primarily using airborne repeat pass fully polarimetric interferometry at L-, and P-band and more recently even at X-band. The experiments demonstrated the potential of this new technology to estimate with high accuracy key forest parameters like forest height, and - using alometric relations - above ground forest biomass. In this paper the actual status of Pol-InSAR forest applications at L-band are reviewed and the potential as well as the limitations of the approaches are discussed by means of results obtained in the frame of validation experiments in very different forest and terrain conditions. Future trends in forest parameter estimation by means of Pol- InSAR are discussed. The majority of the next generation spaceborne polarimetric SAR systems are designed to operate in a repeat pass interferometric mode. This allows the application of Pol- / InSAR parameter inversion tech-niques for forest parameter mapping and characterisation. However, the most critical problem for a success-ful implementation of Pol-InSAR parameter inversion techniques is temporal decorrelation. Under this light, the quantification and correction of temporal decorre-lation becomes an important and actual issue. Regarding forest vegetation, temporal decorrelation is related primarily to the temporal stability of the spatial distribution of the scatterers within the resolution cell and is superimposed to the volume decorrelation contribution that is related to the vertical distribution of scatterers within the resolution cell and contains the information about vertical structure. Similar to any other system induced decorrelation contribution, temporal decorrelation reduces the per-formance of a Pol-InSAR configuration by biasing the (absolute value of the) volume decorrelation contribution and increasing at the same time the standard de-viation of the InSAR phase - for the same number of looks. This leads - if not accounted - to biased esti-mates characterised by a higher variance. We assess and quantify the amount of temporal decorrelation as a function of temporal baseline based on experimental data ac-quired in the frame of several space and airborne campaigns at L-band and discuss the potential to account for in the inversion methodology.