Refine
Year of publication
Document Type
- Article (41)
- Other (2)
- Conference Proceeding (1)
- Doctoral Thesis (1)
- Habilitation Thesis (1)
- Postprint (1)
- Review (1)
Keywords
A volcanic eruption is usually preceded by seismic precursors, but their interpretation and use for forecasting the eruption onset time remain a challenge. A part of the eruptive processes in open conduits of volcanoes may be similar to those encountered in geysers. Since geysers erupt more often, they are useful sites for testing new forecasting methods. We tested the application of Permutation Entropy (PE) as a robust method to assess the complexity in seismic recordings of the Strokkur geyser, Iceland. Strokkur features several minute-long eruptive cycles, enabling us to verify in 63 recorded cycles whether PE behaves consistently from one eruption to the next one. We performed synthetic tests to understand the effect of different parameter settings in the PE calculation. Our application to Strokkur shows a distinct, repeating PE pattern consistent with previously identified phases in the eruptive cycle. We find a systematic increase in PE within the last 15 s before the eruption, indicating that an eruption will occur. We quantified the predictive power of PE, showing that PE performs better than seismic signal strength or quiescence when it comes to forecasting eruptions.
A review of source models to further the understanding of the seismicity of the Groningen field
(2022)
The occurrence of felt earthquakes due to gas production in Groningen has initiated numerous studies and model attempts to understand and quantify induced seismicity in this region. The whole bandwidth of available models spans the range from fully deterministic models to purely empirical and stochastic models. In this article, we summarise the most important model approaches, describing their main achievements and limitations. In addition, we discuss remaining open questions and potential future directions of development.
The Coulomb failure stress (CFS) criterion is the most commonly used method for predicting spatial distributions of aftershocks following large earthquakes. However, large uncertainties are always associated with the calculation of Coulomb stress change. The uncertainties mainly arise due to nonunique slip inversions and unknown receiver faults; especially for the latter, results are highly dependent on the choice of the assumed receiver mechanism. Based on binary tests (aftershocks yes/no), recent studies suggest that alternative stress quantities, a distance-slip probabilistic model as well as deep neural network (DNN) approaches, all are superior to CFS with predefined receiver mechanism. To challenge this conclusion, which might have large implications, we use 289 slip inversions from SRCMOD database to calculate more realistic CFS values for a layered half-space and variable receiver mechanisms. We also analyze the effect of the magnitude cutoff, grid size variation, and aftershock duration to verify the use of receiver operating characteristic (ROC) analysis for the ranking of stress metrics. The observations suggest that introducing a layered half-space does not improve the stress maps and ROC curves. However, results significantly improve for larger aftershocks and shorter time periods but without changing the ranking. We also go beyond binary testing and apply alternative statistics to test the ability to estimate aftershock numbers, which confirm that simple stress metrics perform better than the classic Coulomb failure stress calculations and are also better than the distance-slip probabilistic model.
Based on an analysis of continuous monitoring of farm animal behavior in the region of the 2016 M6.6 Norcia earthquake in Italy, Wikelski et al., 2020; (Seismol Res Lett, 89, 2020, 1238) conclude that animal activity can be anticipated with subsequent seismic activity and that this finding might help to design a "short-term earthquake forecasting method." We show that this result is based on an incomplete analysis and misleading interpretations. Applying state-of-the-art methods of statistics, we demonstrate that the proposed anticipatory patterns cannot be distinguished from random patterns, and consequently, the observed anomalies in animal activity do not have any forecasting power.
Groningen is the largest onshore gas field under production in Europe. The pressure depletion of the gas field started in 1963. In 1991, the first induced micro-earthquakes have been located at reservoir level with increasing rates in the following decades. Most of these events are of magnitude less than 2.0 and cannot be felt. However, maximum observed magnitudes continuously increased over the years until the largest, significant event with ML=3.6 was recorded in 2014, which finally led to the decision to reduce the production. This causal sequence displays the crucial role of understanding and modeling the relation between production and induced seismicity for economic planing and hazard assessment. Here we test whether the induced seismicity related to gas exploration can be modeled by the statistical response of fault networks with rate-and-state-dependent frictional behavior. We use the long and complete local seismic catalog and additionally detailed information on production-induced changes at the reservoir level to test different seismicity models. Both the changes of the fluid pressure and of the reservoir compaction are tested as input to approximate the Coulomb stress changes. We find that the rate-and-state model with a constant tectonic background seismicity rate can reproduce the observed long delay of the seismicity onset. In contrast, so-called Coulomb failure models with instantaneous earthquake nucleation need to assume that all faults are initially far from a critical state of stress to explain the delay. Our rate-and-state model based on the fluid pore pressure fits the spatiotemporal pattern of the seismicity best, where the fit further improves by taking the fault density and orientation into account. Despite its simplicity with only three free parameters, the rate-and-state model can reproduce the main statistical features of the observed activity.
Geysers are hot springs whose frequency of water eruptions remain poorly understood. We set up a local broadband seismic network for 1 year at Strokkur geyser, Iceland, and developed an unprecedented catalog of 73,466 eruptions. We detected 50,135 single eruptions but find that the geyser is also characterized by sets of up to six eruptions in quick succession. The number of single to sextuple eruptions exponentially decreased, while the mean waiting time after an eruption linearly increased (3.7 to 16.4 min). While secondary eruptions within double to sextuple eruptions have a smaller mean seismic amplitude, the amplitude of the first eruption is comparable for all eruption types. We statistically model the eruption frequency assuming discharges proportional to the eruption multiplicity and a constant probability for subsequent events within a multituple eruption. The waiting time after an eruption is predictable but not the type or amplitude of the next one. <br /> Plain Language Summary Geysers are springs that often erupt in hot water fountains. They erupt more often than volcanoes but are quite similar. Nevertheless, it is poorly understood how often volcanoes and also geysers erupt. We created a list of 73,466 eruption times of Strokkur geyser, Iceland, from 1 year of seismic data. The geyser erupted one to six times in quick succession. We found 50,135 single eruptions but only 1 sextuple eruption, while the mean waiting time increased from 3.7 min after single eruptions to 16.4 min after sextuple eruptions. Mean amplitudes of each eruption type were higher for single eruptions, but all first eruptions in a succession were similar in height. Assuming a constant heat inflow at depth, we can predict the waiting time after an eruption but not the type or amplitude of the next one.
The Seismic Hazard Inferred from Tectonics based on the Global Strain Rate Map (SHIFT_GSRM) earthquake forecast was designed to provide high-resolution estimates of global shallow seismicity to be used in seismic hazard assessment. This model combines geodetic strain rates with global earthquake parameters to characterize long-term rates of seismic moment and earthquake activity. Although SHIFT_GSRM properly computes seismicity rates in seismically active continental regions, it underestimates earthquake rates in subduction zones by an average factor of approximately 3. We present a complementary method to SHIFT_GSRM to more accurately forecast earthquake rates in 37 subduction segments, based on the conservation of moment principle and the use of regional interface seismicity parameters, such as subduction dip angles, corner magnitudes, and coupled seismogenic thicknesses. In seven progressive steps, we find that SHIFT_GSRM earthquake-rate underpredictions are mainly due to the utilization of a global probability function of seismic moment release that poorly captures the great variability among subduction megathrust interfaces. Retrospective test results show that the forecast is consistent with the observations during the 1 January 1977 to 31 December 2014 period. Moreover, successful pseudoprospective evaluations for the 1 January 2015 to 31 December 2018 period demonstrate the power of the regionalized earthquake model to properly estimate subduction-zone seismicity.
The Gutenberg-Richter relation for earthquake magnitudes is the most famous empirical law in seismology. It states that the frequency of earthquake magnitudes follows an exponential distribution; this has been found to be a robust feature of seismicity above the completeness magnitude, and it is independent of whether global, regional, or local seismicity is analyzed. However, the exponent b of the distribution varies significantly in space and time, which is important for process understanding and seismic hazard assessment; this is particularly true because of the fact that the Gutenberg-Richter b-value acts as a proxy for the stress state and quantifies the ratio of large-to-small earthquakes. In our work, we focus on the automatic detection of statistically significant temporal changes of the b-value in seismicity data. In our approach, we use Bayes factors for model selection and estimate multiple change-points of the frequency-magnitude distribution in time. The method is first applied to synthetic data, showing its capability to detect change-points as function of the size of the sample and the b-value contrast. Finally, we apply this approach to examples of observational data sets for which b-value changes have previously been stated. Our analysis of foreshock and after-shock sequences related to mainshocks, as well as earthquake swarms, shows that only a portion of the b-value changes is statistically significant.
We observe remarkably periodic patterns of seismicity rates and magnitudes at the Fimbul Ice Shelf, East Antarctica, correlating with the cycles of the ocean tide. Our analysis covers 19 years of continuous seismic recordings from Antarctic broadband stations. Seismicity commences abruptly during austral summer 2011 at a location near the ocean front in a shallow water region. Dozens of highly repetitive events occur in semi-diurnal cycles, with magnitudes and rates fluctuating steadily with the tide. In contrast to the common unpredictability of earthquake magnitudes, the event magnitudes show deterministic trends within single cycles and strong correlations with spring tides and tide height. The events occur quasi-periodically and the highly constrained event sources migrate landwards during rising tide. We show that a simple, mechanical model can explain most of the observations. Our model assumes stick-slip motion on a patch of grounded ice shelf, which is forced by the variations of the ocean-tide height and ice flow. The well fitted observations give new insights into the general process of frictional triggering of earthquakes, while providing independent evidence of variations in ice shelf thickness and grounding.
Earthquake rates are driven by tectonic stress buildup, earthquake-induced stress changes, and transient aseismic processes. Although the origin of the first two sources is known, transient aseismic processes are more difficult to detect. However, the knowledge of the associated changes of the earthquake activity is of great interest, because it might help identify natural aseismic deformation patterns such as slow-slip events, as well as the occurrence of induced seismicity related to human activities. For this goal, we develop a Bayesian approach to identify change-points in seismicity data automatically. Using the Bayes factor, we select a suitable model, estimate possible change-points, and we additionally use a likelihood ratio test to calculate the significance of the change of the intensity. The approach is extended to spatiotemporal data to detect the area in which the changes occur. The method is first applied to synthetic data showing its capability to detect real change-points. Finally, we apply this approach to observational data from Oklahoma and observe statistical significant changes of seismicity in space and time.
In public perception, abnormal animal behavior is widely assumed to be a potential earthquake precursor, in strong contrast to the viewpoint in natural sciences. Proponents of earthquake prediction via animals claim that animals feel and react abnormally to small changes in environmental and physico-chemical parameters related to the earthquake preparation process. In seismology, however, observational evidence for changes of physical parameters before earthquakes is very weak. In this study, we reviewed 180 publications regarding abnormal animal behavior before earthquakes and analyze and discuss them with respect to (1) magnitude-distance relations, (2) foreshock activity, and (3) the quality and length of the published observations. More than 700 records of claimed animal precursors related to 160 earthquakes are reviewed with unusual behavior of more than 130 species. The precursor time ranges from months to seconds prior to the earthquakes, and the distances from a few to hundreds of kilometers. However, only 14 time series were published, whereas all other records are single observations. The time series are often short (the longest is 1 yr), or only small excerpts of the full data set are shown. The probability density of foreshocks and the occurrence of animal precursors are strikingly similar, suggesting that at least parts of the reported animal precursors are in fact related to foreshocks. Another major difficulty for a systematic and statistical analysis is the high diversity of data, which are often only anecdotal and retrospective. The study clearly demonstrates strong weaknesses or even deficits in many of the published reports on possible abnormal animal behavior. To improve the research on precursors, we suggest a scheme of yes and no questions to be assessed to ensure the quality of such claims.
Aseismic transient driving the swarm-like seismic sequence in the Pollino range, Southern Italy
(2015)
Tectonic earthquake swarms challenge our understanding of earthquake processes since it is difficult to link observations to the underlying physical mechanisms and to assess the hazard they pose. Transient forcing is thought to initiate and drive the spatio-temporal release of energy during swarms. The nature of the transient forcing may vary across sequences and range from aseismic creeping or transient slip to diffusion of pore pressure pulses to fluid redistribution and migration within the seismogenic crust. Distinguishing between such forcing mechanisms may be critical to reduce epistemic uncertainties in the assessment of hazard due to seismic swarms, because it can provide information on the frequency-magnitude distribution of the earthquakes (often deviating from the assumed Gutenberg-Richter relation) and on the expected source parameters influencing the ground motion (for example the stress drop). Here we study the ongoing Pollino range (Southern Italy) seismic swarm, a long-lasting seismic sequence with more than five thousand events recorded and located since October 2010. The two largest shocks (magnitude M-w = 4.2 and M-w = 5.1) are among the largest earthquakes ever recorded in an area which represents a seismic gap in the Italian historical earthquake catalogue. We investigate the geometrical, mechanical and statistical characteristics of the largest earthquakes and of the entire swarm. We calculate the focal mechanisms of the M-l > 3 events in the sequence and the transfer of Coulomb stress on nearby known faults and analyse the statistics of the earthquake catalogue. We find that only 25 per cent of the earthquakes in the sequence can be explained as aftershocks, and the remaining 75 per cent may be attributed to a transient forcing. The b-values change in time throughout the sequence, with low b-values correlated with the period of highest rate of activity and with the occurrence of the largest shock. In the light of recent studies on the palaeoseismic and historical activity in the Pollino area, we identify two scenarios consistent with the observations and our analysis: This and past seismic swarms may have been 'passive' features, with small fault patches failing on largely locked faults, or may have been accompanied by an 'active', largely aseismic, release of a large portion of the accumulated tectonic strain. Those scenarios have very different implications for the seismic hazard of the area.
In this study, we analyze acoustic emission (AE) data recorded at the Morsleben salt mine, Germany, to assess the catalog completeness, which plays an important role in any seismicity analysis. We introduce the new concept of a magnitude completeness interval consisting of a maximum magnitude of completeness (M-c(max)) in addition to the well-known minimum magnitude of completeness. This is required to describe the completeness of the catalog, both for the smallest events (for which the detection performance may be low) and for the largest ones (which may be missed because of sensors saturation). We suggest a method to compute the maximum magnitude of completeness and calculate it for a spatial grid based on (1) the prior estimation of saturation magnitude at each sensor, (2) the correction of the detection probability function at each sensor, including a drop in the detection performance when it saturates, and (3) the combination of detection probabilities of all sensors to obtain the network detection performance. The method is tested using about 130,000 AE events recorded in a period of five weeks, with sources confined within a small depth interval, and an example of the spatial distribution of M-c(max) is derived. The comparison between the spatial distribution of M-c(max) and of the maximum possible magnitude (M-max), which is here derived using a recently introduced Bayesian approach, indicates that M-max exceeds M-c(max) in some parts of the mine. This suggests that some large and important events may be missed in the catalog, which could lead to a bias in the hazard evaluation.
Stress drop is a key factor in earthquake mechanics and engineering seismology. However, stress drop calculations based on fault slip can be significantly biased, particularly due to subjectively determined smoothing conditions in the traditional least-square slip inversion. In this study, we introduce a mechanically constrained Bayesian approach to simultaneously invert for fault slip and stress drop based on geodetic measurements. A Gaussian distribution for stress drop is implemented in the inversion as a prior. We have done several synthetic tests to evaluate the stability and reliability of the inversion approach, considering different fault discretization, fault geometries, utilized datasets, and variability of the slip direction, respectively. We finally apply the approach to the 2010 M8.8 Maule earthquake and invert for the coseismic slip and stress drop simultaneously. Two fault geometries from the literature are tested. Our results indicate that the derived slip models based on both fault geometries are similar, showing major slip north of the hypocenter and relatively weak slip in the south, as indicated in the slip models of other studies. The derived mean stress drop is 5-6 MPa, which is close to the stress drop of similar to 7 MPa that was independently determined according to force balance in this region Luttrell et al. (J Geophys Res, 2011). These findings indicate that stress drop values can be consistently extracted from geodetic data.
Earthquakes occurring close to hydrocarbon fields under production are often under critical view of being induced or triggered. However, clear and testable rules to discriminate the different events have rarely been developed and tested. The unresolved scientific problem may lead to lengthy public disputes with unpredictable impact on the local acceptance of the exploitation and field operations. We propose a quantitative approach to discriminate induced, triggered, and natural earthquakes, which is based on testable input parameters. Maxima of occurrence probabilities are compared for the cases under question, and a single probability of being triggered or induced is reported. The uncertainties of earthquake location and other input parameters are considered in terms of the integration over probability density functions. The probability that events have been human triggered/induced is derived from the modeling of Coulomb stress changes and a rate and state-dependent seismicity model. In our case a 3-D boundary element method has been adapted for the nuclei of strain approach to estimate the stress changes outside the reservoir, which are related to pore pressure changes in the field formation. The predicted rate of natural earthquakes is either derived from the background seismicity or, in case of rare events, from an estimate of the tectonic stress rate. Instrumentally derived seismological information on the event location, source mechanism, and the size of the rupture plane is of advantage for the method. If the rupture plane has been estimated, the discrimination between induced or only triggered events is theoretically possible if probability functions are convolved with a rupture fault filter. We apply the approach to three recent main shock events: (1) the M-w 4.3 Ekofisk 2001, North Sea, earthquake close to the Ekofisk oil field; (2) the M-w 4.4 Rotenburg 2004, Northern Germany, earthquake in the vicinity of the Sohlingen gas field; and (3) the M-w 6.1 Emilia 2012, Northern Italy, earthquake in the vicinity of a hydrocarbon reservoir. The three test cases cover the complete range of possible causes: clearly human induced, not even human triggered, and a third case in between both extremes.
Identification and characterization of growing large-scale en-echelon fractures in a salt mine
(2014)
The spatiotemporal seismicity of acoustic emission (AE) events recorded in the Morsleben salt mine is investigated. Almost a year after backfilling of the cavities from 2003, microevents are distributed with distinctive stripe shapes above cavities at different depth levels. The physical forces driving the creation of these stripes are still unknown. This study aims to find the active stripes and track fracture developments over time by combining two different temporal and spatial clustering techniques into a single methodological approach. Anomalous seismicity parameters values like sharp b-value changes for two active stripes are good indicators to explain possible stress accumulation at the stripe tips. We identify the formation of two new seismicity stripes and show that the AE activities in active clusters are migrated mostly unidirectional to eastward and upward. This indicates that the growth of underlying macrofractures is controlled by the gradient of extensional stress. Studying size distribution characteristic in terms of frequency-magnitude distribution and b-value in active phase and phase with constant seismicity rate show that deviations from the Gutenberg-Richter power law can be explained by the inclusion of different activity phases: (1) the inactive period before the formation of macrofractures, which is characterized by a deficit of larger events (higher b-values) and (2) the period of fracture growth characterized by the occurrence of larger events (smaller b-values).
Earthquake catalogs are probably the most informative data source about spatiotemporal seismicity evolution. The catalog quality in one of the most active seismogenic zones in the world, Japan, is excellent, although changes in quality arising, for example, from an evolving network are clearly present. Here, we seek the best estimate for the largest expected earthquake in a given future time interval from a combination of historic and instrumental earthquake catalogs. We extend the technique introduced by Zoller et al. (2013) to estimate the maximum magnitude in a time window of length T-f for earthquake catalogs with varying level of completeness. In particular, we consider the case in which two types of catalogs are available: a historic catalog and an instrumental catalog. This leads to competing interests with respect to the estimation of the two parameters from the Gutenberg-Richter law, the b-value and the event rate lambda above a given lower-magnitude threshold (the a-value). The b-value is estimated most precisely from the frequently occurring small earthquakes; however, the tendency of small events to cluster in aftershocks, swarms, etc. violates the assumption of a Poisson process that is used for the estimation of lambda. We suggest addressing conflict by estimating b solely from instrumental seismicity and using large magnitude events from historic catalogs for the earthquake rate estimation. Applying the method to Japan, there is a probability of about 20% that the maximum expected magnitude during any future time interval of length T-f = 30 years is m >= 9.0. Studies of different subregions in Japan indicates high probabilities for M 8 earthquakes along the Tohoku arc and relatively low probabilities in the Tokai, Tonankai, and Nankai region. Finally, for scenarios related to long-time horizons and high-confidence levels, the maximum expected magnitude will be around 10.
Due to large uncertainties and non-uniqueness in fault slip inversion, the investigation of stress coupling based on the direct comparison of independent slip inversions, for example, between the coseismic slip distribution and the interseismic slip deficit, may lead to ambiguous conclusions. In this study, we therefore adopt the stress-constrained joint inversion in the Bayesian approach of Wang et al., and implement the physical hypothesis of stress coupling as a prior. We test the hypothesis that interseismic locking is coupled with the coseismic rupture, and the early post-seismic deformation is a stress relaxation process in response to the coseismic stress perturbation. We characterize the role of stress coupling in the seismic cycle by evaluating the efficiency of the model to explain the available data. Taking the 2004 M6 Parkfield earthquake as a study case, we find that the stress coupling hypothesis is in agreement with the data. The coseismic rupture zone is found to be strongly locked during the interseismic phase and the post-seismic slip zone is indicated to be weakly creeping. The post-seismic deformation plays an important role to rebuild stress in the coseismic rupture zone. Based on our results for the stress accumulation during both inter- and post-seismic phase in the coseismic rupture zone, together with the coseismic stress drop, we estimate a recurrence time of M6 earthquake in Parkfield around 23-41 yr, suggesting that the duration of 38 yr between the two recent M6 events in Parkfield is not a surprise.
Stress drop is a key factor in earthquake mechanics and engineering seismology. However, stress drop calculations based on fault slip can be significantly biased, particularly due to subjectively determined smoothing conditions in the traditional least-square slip inversion. In this study, we introduce a mechanically constrained Bayesian approach to simultaneously invert for fault slip and stress drop based on geodetic measurements. A Gaussian distribution for stress drop is implemented in the inversion as a prior. We have done several synthetic tests to evaluate the stability and reliability of the inversion approach, considering different fault discretization, fault geometries, utilized datasets, and variability of the slip direction, respectively. We finally apply the approach to the 2010 M8.8 Maule earthquake and invert for the coseismic slip and stress drop simultaneously. Two fault geometries from the literature are tested. Our results indicate that the derived slip models based on both fault geometries are similar, showing major slip north of the hypocenter and relatively weak slip in the south, as indicated in the slip models of other studies. The derived mean stress drop is 5-6 MPa, which is close to the stress drop of similar to 7 MPa that was independently determined according to force balance in this region Luttrell et al. (J Geophys Res, 2011). These findings indicate that stress drop values can be consistently extracted from geodetic data.