Downsampling InSAR measurements is important for efficient geophysical modeling, yet it's vital not to eliminate meaningful signals in the process. How can we achieve this balance? One effective strategy is Quadtree downsampling! Through two courses, "InSAR Processing and Theory with GMTSAR" (July 2019) and "InSAR Processing & Time-Series Analysis for Geophysical Applications: ISCE, ARIA-Tools, and MintPy" (August 2021), I gained many useful skills and knowledge about InSAR techniques, including Quadtree downsampling, among others. These courses were instrumental in my learning journey. I extend my thanks to all the instructors and to UNAVCO (now EarthScope) for hosting these enlightening courses!
(Left [top on mobile]) Inter-seismic InSAR LOS velocity image for the creeping Hayward Fault in the Bay Area, California, was obtained. I processed this InSAR data following the tutorial for MintPy + ARIA Tools (a Jupyter notebook provided by Yunjun Zhang et al.; Link). Small circles indicate downsampled data points obtained using the Pyrocko QuadTree Module (Link). (Right [bottom on mobile]) Examples that show how different thresholds of variance in InSAR measurements affect the QuadTree downsampling.
A plot of the GNSS daily position time series for station P572 is presented (Kim et al., 2021a). It's crucial to apply proper corrections for artificial steps; otherwise, your deformation models might inaccurately indicate significant 'strain rates,' which, in reality, does NOT exist! This critical piece of advice was imparted by Professor Bill Hammond of the University of Nevada, Reno.
I obtained fault orientations for vertical strike-slip faults using equations described by Holt and Haines (1993; https://doi.org/10.1029/92TC00658). Comparing these orientations with focal mechanism solutions in southern California revealed that the discrepancies in strikes were less than 1 degree (Kim et al., 2021). Major faults are highlighted in red (Jennings, 1994). My PhD advisor Professor William E. Holt taught me this cool technique!
My first deformation model for Italy and surrounding regions! The red and black crosses represent the principal axes of the horizontal strain rate field (red for maximum stretching directions; black for maximum shortening directions). The background illustrates dilatational strain rates. It's noteworthy that focal mechanism solutions, which are NOT used in this deformation model, can independently offer valuable insights into the orientations of the strain rate field. See how closely these beach balls align with the deformation model! I conducted this research during my senior undergraduate year(2016-2017), under the guidance of my advisors Professor William E. Holt, Professor Dan M. Davis., and Dr. Alireza Bahadori.