Piyush Agram

Improving InSAR geodesy using Global Atmospheric Models

Spatial and temporal variations of pressure, temperature, and water vapor content in the atmosphere introduce significant confounding delays in interferometric synthetic aperture radar (InSAR) observations of ground deformation and bias estimates of regional strain rates. Producing robust estimates of tropospheric delays remains one of the key challenges in increasing the accuracy of ground deformation measurements using InSAR. Recent studies revealed the efficiency of global atmospheric reanalysis to mitigate the impact of tropospheric delays, motivating further exploration of their potential. Here we explore the effectiveness of these models in several geographic and tectonic settings on both single interferograms and time series analysis products. Both hydrostatic and wet contributions to the phase delay are important to account for. We validate these path delay corrections by comparing with estimates of vertically integrated atmospheric water vapor content derived from the passive multispectral imager Medium-Resolution Imaging Spectrometer, onboard the Envisat satellite. Generally, the performance of the prediction depends on the vigor of atmospheric turbulence. We discuss (1) how separating atmospheric and orbital contributions allows one to better measure long-wavelength deformation and (2) how atmospheric delays affect measurements of surface deformation following earthquakes, and (3) how such a method allows us to reduce biases in multiyear strain rate estimates by reducing the influence of unevenly sampled seasonal oscillations of the tropospheric delay.

Multiscale InSAR Time Series (MInTS) analysis of surface deformation

[1] We present a new approach to extracting spatially and temporally continuous ground deformation fields from interferometric synthetic aperture radar (InSAR) data. We focus on unwrapped interferograms from a single viewing geometry, estimating ground deformation along the line-of-sight. Our approach is based on a wavelet decomposition in space and a general parametrization in time. We refer to this approach as MInTS (Multiscale InSAR Time Series). The wavelet decomposition efficiently deals with commonly seen spatial covariances in repeat-pass InSAR measurements, since the coefficients of the wavelets are essentially spatially uncorrelated. Our time-dependent parametrization is capable of capturing both recognized and unrecognized processes, and is not arbitrarily tied to the times of the SAR acquisitions. We estimate deformation in the wavelet-domain, using a cross-validated, regularized least squares inversion. We include a model-resolution-based regularization, in order to more heavily damp the model during periods of sparse SAR acquisitions, compared to during times of dense acquisitions. To illustrate the application of MInTS, we consider a catalog of 92 ERS and Envisat interferograms, spanning 16 years, in the Long Valley caldera, CA, region. MInTS analysis captures the ground deformation with high spatial density over the Long Valley region.

Aseismic slip and seismogenic coupling along the central San Andreas Fault

We use high-resolution Synthetic Aperture Radar- and GPS-derived observations of surface displacements to derive the first probabilistic estimates of fault coupling along the creeping section of the San Andreas Fault, in between the terminations of the 1857 and 1906 magnitude 7.9 earthquakes. Using a fully Bayesian approach enables unequaled resolution and allows us to infer a high probability of significant fault locking along the creeping section. The inferred discreet locked asperities are consistent with evidence for magnitude 6+ earthquakes over the past century in this area and may be associated with the initiation phase of the 1857 earthquake. As creeping segments may be related to the initiation and termination of seismic ruptures, such distribution of locked and creeping asperities highlights the central role of the creeping section on the occurrence of major earthquakes along the San Andreas Fault.

New Radar Interferometric Time Series Analysis Toolbox Released

Interferometric synthetic aperture radar (InSAR) has become an important geodetic tool for measuring deformation of Earth’s surface due to various geophysical phenomena, including slip on earthquake faults, subsurface migration of magma, slow‐moving landslides, movement of shallow crustal fluids (e.g., water and oil), and glacier flow. Airborne and spaceborne synthetic aperture radar (SAR) instruments transmit microwaves toward Earth’s surface and detect the returning reflected waves. The phase of the returned wave depends on the distance between the satellite and the surface, but it is also altered by atmospheric and other effects. InSAR provides measurements of surface deformation by combining amplitude and phase information from two SAR images of the same location taken at different times to create an interferogram. Several existing open‐source analysis tools [Rosen et al., 2004; Rosen et al., 2011; Kampes et al., 2003 ; Sandwell et al., 2011] enable scientists to exploit observations from radar satellites acquired at two different epochs to produce a surface displacement map.

Change of apparent segmentation of the San Andreas fault around Parkfield from space geodetic observations across multiple periods

Sequences of earthquakes are commonly represented as a succession of periods of interseismic stress accumulation followed by coseismic and postseismic phases of stress release. Because the recurrence time of large earthquakes is often greater than the available span of space geodetic data, it has been challenging to monitor the evolution of interseismic loading in its entire duration. Here we analyze large data sets of surface deformation at different key episodes around the Cholame, Parkfield and creeping segments of the San Andreas Fault that show evidence of significant deceleration of fault slip during the interseismic period. We compare the average fault slip rates before and after the 2004 Mw6 Parkfield earthquake, in the 1986–2004 and 2006–2012 periods, respectively, avoiding 2 years of postseismic deformation after 2004. Using a combination of GPS data from the Plate Boundary Observatory, the Southern California Earthquake Center Crustal Motion Map and the Bay Area Velocity Unification networks and interferometric synthetic aperture radar from the Advanced Land Observing Satellite (ALOS) and Envisat satellites, we show that the area of coupling at the transition between the Parkfield and Cholame segments appears larger later in the interseismic period than it does earlier on. While strong plate coupling is uniform across the Parkfield and Cholame segments in the 1986–2004 period, creep occurs south of the 2004 epicenter after 2006, making segmentation of the San Andreas Fault south of Parkfield more clearly apparent. These observations indicate that analyses of surface deformation late in the earthquake cycle may overestimate the area of plate coupling. A fault surface creeping much below plate rate may in some case be a region that does not promote earthquake nucleation but rather just be at a slower stage of its evolution. Our analysis also shows signs of large variation of slip velocity above and below plate rate in the creeping segment indicating that cycles of weakening and hardening can also be at play in dominantly aseismic areas.

High quality InSAR data linked to seasonal change in hydraulic head for an agricultural area in the San Luis Valley, Colorado

In the San Luis Valley (SLV), Colorado legislation passed in 2004 requires that hydraulic head levels in the confined aquifer system stay within the range experienced in the years 1978–2000. While some measurements of hydraulic head exist, greater spatial and temporal sampling would be very valuable in understanding the behavior of the system. Interferometric synthetic aperture radar (InSAR) data provide fine spatial resolution measurements of Earth surface deformation, which can be related to hydraulic head change in the confined aquifer system. However, change in cm-scale crop structure with time leads to signal decorrelation, resulting in low quality data. Here we apply small baseline subset (SBAS) analysis to InSAR data collected from 1992 to 2001. We are able to show high levels of correlation, denoting high quality data, in areas between the center pivot irrigation circles, where the lack of water results in little surface vegetation. At three well locations we see a seasonal variation in the InSAR data that mimics the hydraulic head data. We use measured values of the elastic skeletal storage coefficient to estimate hydraulic head from the InSAR data. In general the magnitude of estimated and measured head agree to within the calculated error. However, the errors are unacceptably large due to both errors in the InSAR data and uncertainty in the measured value of the elastic skeletal storage coefficient. We conclude that InSAR is capturing the seasonal head variation, but that further research is required to obtain accurate hydraulic head estimates from the InSAR deformation measurements.

A noise model for InSAR time series

Interferometric synthetic aperture radar (InSAR) time series methods estimate the spatiotemporal evolution of surface deformation by incorporating information from multiple SAR interferograms. While various models have been developed to describe the interferometric phase and correlation statistics in individual interferograms, efforts to model the generalized covariance matrix that is directly applicable to joint analysis of networks of interferograms have been limited in scope. In this work, we build on existing decorrelation and atmospheric phase screen models and develop a covariance model for interferometric phase noise over space and time. We present arguments to show that the exploitation of the full 3-D covariance structure within conventional time series inversion techniques is computationally challenging. However, the presented covariance model can aid in designing new inversion techniques that can at least mitigate the impact of spatial correlated nature of InSAR observations.

Detecting transient signals in geodetic time series using sparse estimation techniques

We present a new method for automatically detecting transient deformation signals from geodetic time series. We cast the detection problem as a least squares procedure where the design matrix corresponds to a highly overcomplete, nonorthogonal dictionary of displacement functions in time that resemble transient signals of various timescales. The addition of a sparsity-inducing regularization term to the cost function limits the total number of dictionary elements needed to reconstruct the signal. Sparsity-inducing regularization enhances interpretability of the resultant time-dependent model by localizing the dominant timescales and onset times of the transient signals. Transient detection can then be performed using convex optimization software where detection sensitivity is dependent on the strength of the applied sparsity-inducing regularization. To assess uncertainties associated with estimation of the dictionary coefficients, we compare solutions with those found through a Bayesian inference approach to sample the full model space for each dictionary element. In addition to providing uncertainty bounds on the coefficients and confirming the optimization results, Bayesian sampling reveals trade-offs between dictionary elements that have nearly equal probability in modeling a transient signal. Thus, we can rigorously assess the probabilities of the occurrence of transient signals and their characteristic temporal evolution. The detection algorithm is applied on several synthetic time series and real observed GPS time series for the Cascadia region. For the latter data set, we incorporate a spatial weighting scheme that self-adjusts to the local network density and filters for spatially coherent signals. The weighting allows for the automatic detection of repeating slow slip events.

A Network-Based Enhanced Spectral Diversity Approach for TOPS Time-Series Analysis

For multitemporal analysis of synthetic aperture radar (SAR) images acquired with a terrain observation by progressive scan (TOPS) mode, all acquisitions from a given satellite track must be coregistered to a reference coordinate system with accuracies better than 0.001 of a pixel (assuming full SAR resolution) in the azimuth direction. Such a high accuracy can be achieved through geometric coregistration, using precise satellite orbits and a digital elevation model, followed by a refinement step using a time-series analysis of coregistration errors. These errors represent the misregistration between all TOPS acquisitions relative to the reference coordinate system. We develop a workflow to estimate the time series of azimuth misregistration using a network-based enhanced spectral diversity (NESD) approach, in order to reduce the impact of temporal decorrelation on coregistration. Example time series of misregistration inferred for five tracks of Sentinel-1 TOPS acquisitions indicates a maximum relative azimuth misregistration of less than 0.01 of the full azimuth resolution between the TOPS acquisitions in the studied areas. Standard deviation of the estimated misregistration time series for different stacks varies from 1.1e-3 to 2e-3 of the azimuth resolution, equivalent to 1.6-2.8 cm orbital uncertainty in the azimuth direction. These values fall within the 1-sigma orbital uncertainty of the Sentinel-1 orbits and imply that orbital uncertainty is most likely the main source of the constant azimuth misregistration between different TOPS acquisitions. We propagate the uncertainty of individual misregistration estimated with ESD to the misregistration time series estimated with NESD and investigate the different challenges for operationalizing NESD.

Aseismic slip and seismogenic coupling in the Marmara Sea: What can we learn from onland geodesy?

Ever since the Mw7.4 Izmit earthquake in 1999, evaluation of seismic hazard associated with the last unbroken segments of the North Anatolian fault is capital. A strong controversy remains over whether Marmara fault segments are locked or are releasing strain aseismically. Using a Bayesian approach, we propose a preliminary probabilistic interseismic model constrained by published GPS data sets. The posterior mean model show that Ganos and Cinarcik segments are locked while creep is detected in the central portion of Marmara fault. Our analysis, however, reveals that creeping segments are associated with large model uncertainties, which mainly results from the sparsity of current geodetic observations. We then discuss how the GPS network can be improved to attain more reliable assessment of interseismic slip rates. With this purpose, we implement a network optimization procedure to identify the most favorable distribution of stations measuring strain accumulation in the Marmara Sea.