Subtle volcanic deformations point to volcanic activities, and monitoring them helps predict eruptions. Today, it is possible to remotely detect volcanic deformation in mm/year scale thanks to advances in interferometric Synthetic Aperture Radar (InSAR). This paper proposes a framework based on a deep learning model to automatically discriminate subtle volcanic deformations from other deformation types in five-year-long InSAR stacks. Models are trained on a synthetic training set. To better understand and improve the models, explainable AI analyses are performed. In initial models, gradient-weighted Class Activation Mapping (Grad-CAM) linked new-found patterns of slope processes and salt lake deformations to false-positive detections. The models are then improved by fine-tuning with a hybrid synthetic-real data, and additional performance is extracted by low-pass spatial filtering of the real test set. T-SNE latent feature visualization confirmed the similarity and shortcomings of the fine-tuning set, highlighting the problem of elevation components in residual tropospheric noise. After fine-tuning, all the volcanic deformations are detected, including the smallest one, Lazufre, deforming 5 mm/year. The first time confirmed deformation of Cerro El Condor is observed, deforming 9.9-17.5 mm/year. Finally, sensitivity analysis uncovered the model’s minimal detectable deformation of 2 mm/year.

With improvement in the processing of synthetic aperture radar interferometry (InSAR) data, the detection of long-term volcanic deformations becomes possible. While deep learning (DL) models are considered black-box models, challenging to debug, the advances in explainable AI (XAI) help understand the model and how it makes decisions. In this paper, the model is trained on synthetic InSAR velocity maps to detect slow, sustained deformations. XAI tools, including Grad-CAM and t-SNE, are utilized for understanding and improving the trained model. Grad-CAM helps identify the slope-induced signal and salt lake patterns responsible for the model’s mis-classifications. T-SNE feature representation visualizations are used to estimate data sets and model class separation ability. Additionally, a sensitivity analysis shows the model performance with different intensity deformation data and uncovers the minimal detectable deformations of 1 cm cumulative deformation over five years.

Interferometric Synthetic Aperture Radar (InSAR)-derived surface displacement time series enable a wide range of applications from urban structural monitoring to geohazard assessment. With systematic data acquisitions becoming the new norm for SAR missions, millions of time series are continuously generated. Machine Learning provides a framework for the efficient mining of such big data. Here, we focus on unsupervised mining of the data via clustering the similar temporal patterns and data-driven displacement signal reconstruction from the InSAR time series. We propose a deep Long Short Term Memory (LSTM) autoencoder model which can exploit temporal relations in contrast to the commonly used shallow learning methods, such as Uniform Manifold Approximation and Projection (UMAP). We also modify the loss function to allow the quantification of uncertainties in the time series data. The two approaches are applied to the Lazufre Volcanic Complex located at the central volcanic zone of the Andes and thereby compared.

We reveal the presence of a peculiar physical signal which compromises the accuracy of Earth surface deformation estimates for distributed scatterers [1]. The observed signal is short-lived and decays with the temporal baseline; however, it is distinct from the stochastic noise attributed to temporal decorrelation. To indicate its nature, this physical effect is referred to as fading signal. Designing a simple approach in the evaluation of distributed scatterers, we reveal a prominent bias in the deformation velocity maps. The bias is the result of propagation of small phase error through the time series. We further discuss the role of the phase estimation algorithms in significant reduction of the bias and put forward the idea of a unified analysis-ready InSAR product for achieving high-precision deformation monitoring.

Interferometric Synthetic Aperture Radar (InSAR) is a powerful tool for monitoring ground deformation associated with earthquakes, volcanoes, landslides, and different anthropogenic activities. The accuracy of the estimated deformation depends on a number of parameters including tropospheric and ionospheric delays, unwrapping errors, phase decorrelation due to changes in scattering behavior and system noise. However, recently an additional source of phase noise has been identified [1], which is strongest in short-interval multi-looked interferograms and, unlike other sources of noise, leads to biased, non-zero loop closure phases. This is problematic for time-series analysis because short-interval interferograms may be the only ones that maintain coherence for some areas. In this study, we explore the characteristics of this phenomenon and propose a mitigation strategy.

This article investigates the presence of a new interferometric signal in multilooked synthetic aperture radar (SAR) interferograms that cannot be attributed to the atmospheric or Earth-surface topography changes. The observed signal is short-lived and decays with the temporal baseline; however, it is distinct from the stochastic noise attributed to temporal decorrelation. The presence of such afading signalintroduces a systematic phase component, particularly in short temporal baseline interferograms. If unattended, it biases the estimation of Earth surface deformation from SAR time series. Here, the contribution of the mentioned phase component is quantitatively assessed. The biasing impact on the deformation-signal retrieval is further evaluated. A quality measure is introduced to allow the prediction of the associated error with the fading signals. Moreover, a practical solution for the mitigation of this physical signal is discussed; special attention is paid to the efficient processing of Big Data from modern SAR missions such as Sentinel-1 and NISAR. Adopting the proposed solution, the deformation bias is shown to decrease significantly. Based on these analyses, we put forward our recommendations for efficient and accurate deformation-signal retrieval from large stacks of multilooked interferograms.

The presence of stochastic and systematic inconsistencies is a concern for the precision and interpretability of Interferometric Synthetic Aperture Radar (InSAR) when distributed scatterers are exploited for SAR time series analysis [1]. Multitemporal phase estimators aim at retrieving a consistent common-master interferometric time series, thereby reducing the effect of stochastic inconsistencies. Exploiting data redundancy the latter estimators are theoretically expected to decrease the susceptibility to systematic inconsistencies as well. In this contribution we seek a computationally efficient quality measure to show the effectiveness of multitemporal phase estimation in the reduction of inconsistencies. Choosing the ensemble coherence as a candidate, we firstly seek a constant false alarm rate detector for initial detection of signal-bearing areas. Furthermore the impact of phase inconsistencies on the ensemble coherence will be brought into attention.

Multitemporal phase estimation aims at the exploitation temporal data redundancy within the SAR time-series to reduce the impact of inherent stochastic and systematic interferometric phase inconsistencies [1] for distributed scatterers (DS). The consistent phase-series estimated as such is further utilized to retrieve the underlying geophysical and atmospheric signals. Therefore, the precision and interpretability of the retrieved physical signals from the DS is governed by the performance of the phase estimators. Different approaches to phase estimation calls for the investigation of their performance. Here we explain the discrepancy among the different approaches in terms of their underlying covariance model and introduce our recently proposed estimator named EMI [2]. Bridging between different approaches via revised mathematical formulation of phase estimation, EMI enhances the estimation precision and computational efficiency of the temporal phase estimation. The performance of different phase estimators is brought into attention via simulation analysis. Using Sentinel-1 time series over the North and East Anatolian Faults, wide area performance analysis is further carried out and will be presented.

The capacity of SAR interferometry to measure surface deformation with accuracy of 1 mm/year or better are well known. However this is typically limited to relative motion at short distance. Thanks to several advances in SAR sensor quality, data availability, orbit determination, processing, and calibration of atmospheric delays it is now possible to achieve that accuracy even across large distances of hundreds of kilometers.In this paper we revise the main contributions to the large scale error, considering available mitigation techniques. We provide a first validation for a processing based on Sentinel-1 data, by comparing our results with GNSS stations.For future SAR's operating at lower frequencies, it is vital to consider ionospheric corrections and likely also the influence of moisture variations in natural scatterers. The choice of processing algorithms, though typically not discussed, can also have a significant effect on the quality of the result.

Signal decorrelation poses a limitation to multipass SAR interferometry. In pursuit of overcoming this limitation to achieve high-precision deformation estimates, different techniques have been developed, with short baseline subset, SqueeSAR, and CAESAR as the overarching schemes. These different analysis approaches raise the question of their efficiency and limitation in phase and consequently deformation estimation. This contribution first addresses this question and then proposes a new estimator with improved performance, called Eigendecomposition-based Maximum-likelihood-estimator of Interferometric phase (EMI). The proposed estimator combines the advantages of the state-of-the-art techniques. Identical to CAESAR, EMI is solved using eigendecomposition; it is therefore computationally efficient and straightforward in implementation. Similar to SqueeSAR, EMI is a maximum-likelihood-estimator; hence, it retains estimation efficiency. The computational and estimation efficiency of EMI renders it as an optimum choice for phase estimation. A further marriage of EMI with the proposed Sequential Estimator by Ansari et al. provides an efficient processing scheme tailored to the analysis of Big InSAR Data. EMI is formulated and verified in relation to the state-of-the-art approaches via mathematical formulation, simulation analysis, and experiments with time series of Sentinel-1 data over the volcanic island of Vulcano, Italy.

Wide-swath satellite missions with short revisit times, such as Sentinel-1 and the planned NISAR and Tandem-L, provide an unprecedented wealth of interferometric time series and open new opportunities for systematic monitoring of the Earth surface. The processing of the emerging Big Data with the state-of-the-art InSAR time series analysis techniques is, however, challenging. This contribution introduces a novel approach, named Sequential Estimator, for efficient estimation of the interferometric phase from the long InSAR time series. The algorithm uses recursive estimation and analysis of the data covariance matrix via division of the data into small batches, followed by compression of the data batches. From each compressed data batch artificial interferograms are formed, resulting in a strong data reduction. This scheme avoids the necessity of re-processing the entire data stack at the face of each new acquisition. It is shown that the proposed estimator introduces negligible degradation compared to the Cramér-Rao Lower Bound. The estimator may therefore be adapted for high-precision Near-Real-Time processing of InSAR and accommodate the conversion of InSAR from an off-line to a monitoring geodetic tool. The performance of the Sequential Estimator is compared to the state-of-the-art techniques via simulations and application to Sentinel-1 data.

Wide-swath synthetic aperture radar (SAR) missions with short revisit times, such as Sentinel-1 and the planned NISAR and Tandem-L, provide an unprecedented wealth of interferometric SAR (InSAR) time series. However, the processing of the emerging Big Data is challenging for state-of-the-art InSAR analysis techniques. This contribution introduces a novel approach, named Sequential Estimator, for efficient estimation of the interferometric phase from long InSAR time series. The algorithm uses recursive estimation and analysis of the data covariance matrix via division of the data into small batches, followed by the compression of the data batches. From each compressed data batch artificial interferograms are formed, resulting in a strong data reduction. Such interferograms are used to link the “older” data batches with the most recent acquisitions and thus to reconstruct the phase time series. This scheme avoids the necessity of reprocessing the entire data stack at the face of each new acquisition. The proposed estimator introduces negligible degradation compared to the Cramer-Rao lower bound under realistic coherence scenarios. The estimator may therefore be adapted for high-precision near-real-time processing of InSAR and accommodate the conversion of InSAR from an offline to a monitoring geodetic tool. The performance of the Sequential Estimator is compared to state-of-the-art techniques via simulations and application to Sentinel-1 data.

The launch of the wide-swath SAR missions with short repeat-pass cycles, such as Sentinel-1, will soon provide an unprecedented large InSAR data archive. Time-series analysis on the rapidly growing data will thus become computationally demanding for a systematic monitoring of earth surface deformation. As the state-of-the-art approach in differential InSAR time-series analysis, the distributed scatterer interferometric (DSI) techniques shall adapt agile processing schemes to deal with the emerging big data; an aspect to which limited attention has been dedicated. In this contribution, a sequential DSI scheme is proposed to address this demand. Based on SAR data reduction, the scheme allows for batch processing of the large data stacks while preserving the performance close to the Cramér-Rao Lower Bound. The performance of theof earth surface deformation. As the state-of-the-art approach in differential InSAR time-series analysis, the distributed scatterer interferometric (DSI) techniques shall adapt agile processing schemes to deal with the emerging big data; an aspect to which limited attention has been dedicated. In this contribution, a sequential DSI scheme is proposed to address this demand. Based on SAR data reduction, the scheme allows for batch processing of the large data stacks while preserving the performance close to the Cramér-Rao Lower Bound. The performance of the proposed sequential estimator is compared to the current DSI algorithms under two contradicting coherence scenarios. The application of the proposed sequential estimator to stacks proposed sequential estimator is compared to the current DSI algorithms under two contradicting coherence scenarios. The application of the proposed sequential estimator to stacks of Sentinel-1 data is ongoing.

A conventional interferometric synthetic aperture radar (SAR) system provides 1-D line-of-sight motion measurements from repeat-pass observations. Two-dimensional motions may be measured by combining two observations from ascending and descending geometries. The third motion component may be retrieved by adding a third geometry and/or by integrating along-track measurements although with much reduced precision compared to the other two components. Several options exist to improve the accuracy of retrieving the third motion component, such as combining left- and right-looking observations or exploiting recently proposed innovative SAR acquisition modes (BiDiSAR and SuperSAR). These options are, however, challenging for future SAR systems based on large reflector antennae, due to lack of capability to electronic beam steering or frequent toggle between left- and right-looking modes. Therefore, in this letter, we assess and compare the realistic acquisition scenarios for a reflector-based SAR in an attempt to optimize the achievable 3-D precision. Investigating the squinted SAR geometry as one of the feasible scenarios, we show that a squint of 13.5° will yield comparable performance to the left-looking acquisition, while further squinting outperforms this or other feasible configurations. As an optimum configuration for 3-D retrieval, the squinted acquisition is further elaborated: the different acquisition plans considering a constellation of two satellites as well as the challenges for data processing are addressed.

Interferometric synthetic aperture radar (InSAR) measurements are merely sensitive to the deformation along the Line of Sight (LOS) direction of the sensor. To improve the geometrical sensitivity and retrieve the three-dimensional deformation, the integration of InSAR from non-coplanar acquisitions as well as fusion with resolution-scale SAR image shift measurements has become a standard approach. Using different statistical measures, we assess and compare the influence of different image acquisition strategies as well as data fusion on the performance of InSAR in 3D deformation retrieval. Integrating nominal InSAR acquisitions, i.e. a set of measurements from ascending and descending tracks acquired from right-looking geometry, a strong correlation between the retrieved 3D parameters in the local vertical-north plane is observable. This correlation is sought to be decreased by non-nominal acquisitions; i.e. left-looking or squinted observations. These acquisition strategies are discussed for consideration in the future L-band mission Tandem-L.

Constraining the deformation analysis to scatterers with high phase coherence, known as persistent scatterers (PS), in the InSAR time series plays a major role in advanced coherent approaches such as Persistent Scatterer Interferometry (PSI). Although crucial in overcoming the shortcomings of conventional InSAR in retrieving the deformation signals, this constraint appears to be too strict and lead to information loss in cases where the scatterers are coherent in merely a short period of the time series. Exploiting such scatterers, referred to as temporal coherent scatterers (TCS), is a necessity in PSI analysis in order to enhance the density of the resulted point clouds and consequently the quality of PSI. In here, statistical properties of the SAR calibrated amplitude time series are exploited to propose a new method for detection of the TCSs and estimation of the coherent intervals. The proposed methodology also allows for estimation of signal to clutter ratio (SCR) as an indication of the phase coherence. The method is evaluated using the simulated SAR stack as well as TerraSAR-X high resolution complex images.