A1408
Title: Flexible estimation of spatial covariance functions from multi-temporal DInSAR data
Authors: Roberta Troilo - Politecnico di Milano (Italy)
Teresa Bortolotti - Politecnico di Milano (Italy) [presenting]
Alessandra Menafoglio - Politecnico di Milano (Italy)
Simone Vantini - Politecnico di Milano (Italy)
Abstract: Sentinel-1 satellites offer extensive synthetic aperture radar data globally, revisiting locations every six days. Leveraging these data, differential interferometric processing techniques yield high-resolution ground displacement images that are accurate to millimeter precision. These insights into evolving ground conditions enable comprehensive monitoring of large areas prone to environmental hazards. Nonetheless, challenges arise when spatial units (e.g., water or vegetated areas) show a non-coherent variability across successive time instants, yielding missing values in single pixels or entire areas consistently missing along time. Although statistical reconstruction of missing data is possible through spatial interpolation (e.g., Kriging), it typically grounds on the second-order structure of the target field, which, besides being unknown, is typically characterized by strong non-stationarities. The challenge of estimating the spatial covariance operator is faced from time-evolving ground displacement images by developing a novel non-parametric methodology which grounds in the theory of functional data analysis. While the non-parametric approach guarantees flexibility to account for the non-stationarity of the field, a Laplacian-type regularization ensures continuity in the reconstructed operator. The methodology is showcased on ground displacement images collected to monitor the Phlegraean Fields, Italy, a region vulnerable to seismic and bradisismic activity.
