Speaker
Description
The polarimetric-interferometric temporal SAR coherence is defined as the normalized correlation between two SAR single-look image (SLC) samples acquired at different times and free of other decorrelation effects such those associated with thermal noise, perpendicular baseline, and processing artifacts [1]. The temporal coherence gauges “how coherent” the signal scattered from the imaged scene within the multi-looked radar resolution cell remains from one acquisition to another. Lower coherence values indicate larger changes compared to higher coherence values. These changes can be geometric (e.g., motion or addition of scatterers) or dielectric (e.g., alteration of moisture content or temperature). Polarization diversity allows the selection of a scattering mechanism from the received signal. As they relate to different properties of the imaged scene, distinct scattering mechanisms within the same resolution cell undergo different changes over time, leading to different temporal coherence levels.
In recent years, some studies have emphasized the significant impact of land cover on temporal coherence and its evolution over time, specifically how multiple temporal coherence samples vary within a time series as the temporal interval increases [2-4]. Generally, scattered signals lose coherency more rapidly as the time gap between acquisitions widens, which is influenced by the structure of the target. Complex and penetrable targets like vegetation tend to decorrelate faster over time compared to bare soil. Such complex targets also likely display greater temporal coherence diversity as a function of polarization, and in relation to radar and observation parameters (e.g., radar wavelengths and incidence angles). Understanding the interplay between the characteristics of the imaged scene (e.g., vegetation) and the polarimetric-interferometric coherence for different time intervals, as well as the opportunities for parameter estimation, is an active area of research in the Bio-GeoSAR community.
This talk aims to offer the community a comprehensive review of the most recent advances in temporal coherence modeling, coupled with potential parameter estimation algorithms grounded in these models. The primary temporal coherence model to be presented in detail is based on the random-motion-over-ground model (RMoG+) [5]. This model factors in the addition of backscatter soil/canopy changes between acquisitions and the dependency on the temporal interval to model coherence time series. We will evaluate the model using available data sets, including time series from Sentinel-1, ALOS-2, and potentially airborne UAVSAR/AirMOSS time-series at L and P bands. This research work is topical considering current and upcoming L-band SAR missions like SAOCOM, NISAR (2024), ROSE-L (2028/2030), and ALOS-4. These missions will capture SLC time-series within a narrow orbital tube, enabling global, recurrent observations of temporal coherence and presenting novel opportunities for land parameter estimation.
[1] Zebker, H. A., and J. Villasenor, "Decorrelation in interferometric radar echoes," in IEEE Transactions on Geoscience and Remote Sensing, vol. 30, no. 5, pp. 950-959, Sept. 1992, doi: 10.1109/36.175330.
[2] Lavalle, M., C. Telli, N. Pierdicca, U. Khati, O. Cartus, and J. Kellndorfer, “Model-based retrieval of forest parameters from Sentinel-1 coherence and backscatter time series,” IEEE Geoscience and Remote Sensing Letters, vol. 20, pp. 1–5, 2023.
[3] Sica, F., A. Pulella, M. Nannini, M. Pinheiro, and P. Rizzoli, “Repeat- pass SAR interferometry for land cover classification: A methodology using Sentinel-1 Short-Time-Series,” Remote Sensing of Environment, vol. 232, pp. 111277, 2019.
[4] Seppi, S.A., C. López-Martinez, M. J. Joseau. "Assessment of L-Band SAOCOM InSAR Coherence and Its Comparison with C-Band: A Case Study over Managed Forests in Argentina." Remote Sensing, vol. 14.22, pp. 5652, 2022.
[5] Lavalle, M., and M. Simard, and S. Hensley, “A temporal decorrelation model for polarimetric radar interferometers,” Geosc. and Rem. Sens., IEEE Trans. on, vol. 50, no. 7, pp. 2880–2888, July 2012.
