A neural encoder for earthquake rate forecasting

Scientific Reports volume 13, Article number: 12350 (2023) Cite this article 481 Accesses 3 Altmetric Metrics details Forecasting the timing of earthquakes is a long-standing challenge. Moreover, it

Scientific Reports volume 13, Article number: 12350 (2023) Cite this article

481 Accesses

3 Altmetric

Metrics details

Forecasting the timing of earthquakes is a long-standing challenge. Moreover, it is still debated how to formulate this problem in a useful manner, or to compare the predictive power of different models. Here, we develop a versatile neural encoder of earthquake catalogs, and apply it to the fundamental problem of earthquake rate prediction, in the spatio-temporal point process framework. The epidemic type aftershock sequence model (ETAS) effectively learns a small number of parameters to constrain the assumed functional forms for the space and time correlations of earthquake sequences (e.g., Omori-Utsu law). Here we introduce learned spatial and temporal embeddings for point process earthquake forecasting models that capture complex correlation structures. We demonstrate the generality of this neural representation as compared with ETAS model using train-test data splits and how it enables the incorporation additional geophysical information. In rate prediction tasks, the generalized model shows \(>4\%\) improvement in information gain per earthquake and the simultaneous learning of anisotropic spatial structures analogous to fault traces. The trained network can be also used to perform short-term prediction tasks, showing similar improvement while providing a 1000-fold reduction in run-time.

The application of machine-learning (ML) for the analysis of seismological data has seen substantial recent progress highlighted by new approaches for the classification and characterization of seismic waveforms1,2, automatic phase picking3, identification of low-magnitude earthquakes4, and catalog declustering5,6. In the development of earthquake catalogs ML approaches have increased the number of detected events by ten folds4 and will possibly reduce travel time dependence for earthquake early warning from the speed of seismic waves to the speed of light7.

However, in earthquake sequence modeling machine learning techniques have yielded limited progress in terms of enabling improved characterizations seismicity patterns8,9. The specific task of forecasting the timing of future seismic events is a longstanding and fundamental challenge both as a basic scientific question and for applied hazard analysis. While in some cases seismic activity features relatively consistent temporal10 or spatial patterns11, the time, location and magnitude of seismicity has remained difficult to predict quantitatively12.

The state-of-the-art approach to this problem in statistical seismology is to represent earthquake sequences as a spatio-temporal point process13,14,15. In this approach, the model is tasked with predicting the instantaneous rate of earthquake occurrence above a certain magnitude, \(\lambda (x, y, t \mid H_{t-})\), where x, y are spatial coordinates (longitude and latitude or map projected coordinates) and t is time. \(H_{t-}\) represents all the information available to the model prior to time t. The time-dependent function \(\lambda\) is the quantitative representation of the intensity of seismic activity, characterizing both the foreshock16,17 and aftershock18 epochs as well as serving as the foundation for seismic hazard assessment19.

The epidemic-type aftershock sequence (ETAS) model13,20 is the most commonly used such model, representing \(\lambda\) as a self-exciting branching process, which assumes a “background rate” of seismicity and a response function, f, whose specific form is chosen such that the long-term statistics of synthetic earthquake catalogs generated from the model reproduce the two widely observed phenomenological distributions of seismicity: (1) the Omori-Utsu law of aftershock rate decay and (2) the Gutenberg-Richter distribution of event magnitudes. There are a few popular choices for the response function21,22,23,24, that share the form of, \(f = \mu (x,y)+ T(t-t_i)S(x-x_i, y-y_i; M_i)\). Here \(\mu\) is called the time-independent “background rate”, T is a temporal kernel featuring a power-law decay consistent with Omori’s law, and S is a spatially decaying kernel22,25. \(x_i, y_i\), and \(t_i\) are the earthquake’s hypocentral location and occurrence time, respectively.

The ETAS model has been used as an effective representation of earthquake rate changes19,26,27,28. However, its applicability has been limited by several factors. First, finding optimal ETAS parameters is a challenging optimization task, because of a broad minima associated with the the space-dependent background seismicity rate and a range of different parameters for the response function can produce similar log-likelihood scores29,30,31,32,33. Second, the classical predetermined forms of f have a limited expressive power and limit the ETAS approach to the consideration of the hypocenters, times, and magnitudes of past moderate-large magnitude earthquakes. Additional relevant data including small magnitude seismicity, tectonic structure, fault locations and earthquake focal mechanisms are typically not modeled, though some attempts have been made to incorporate them19,21,34,35.

A sketch of the model architecture incorporating information up to time t.

Here we propose the FERN (Forecasting Earthquake Rates with Neural networks) encoder-decoder neural based model to generalize beyond the ETAS constraints. Conceptually, the input is encoded by a neural network to generate a latent representation of the tectonic state, which is then passed to a decoder network (Fig. 1). This design has two specific advantages: first, it naturally allows to incorporate different data sources and modalities, which can be added to the model with source-specific encoders. Second, the same encoded state can be used as input to several prediction heads (“decoders”), which can be used to for different prediction tasks.

This approach matches the performance of the state-of-the-art ETAS model in rate prediction when trained on identical data sets and that the FERN model exhibits increased accuracy when supplied with earthquakes of magnitude smaller than the completeness magnitude threshold of the catalog. We also show how the trained encoders can be used to solve a different prediction problem, a short-term forecast of the number of events in a 24-h period. In this task, the FERN model outperform the ETAS model while requiring 4-5 orders of magnitude less compute time. We do not provide any uncertainty estimates based either on either data error propagation or varying model architecture.

We use three encoders (Fig. 1) to capture different aspects of the seismicity patterns. The recent earthquakes encoder is a direct generalization of the ETAS response function f, replacing the human-engineered functional form of f by a more general neural network. It is intended to capture short-term seismic activity. The long term seismicity encoder learns long-term spatio-temporal seismic patterns by counting earthquake events in varying temporal spans, ranging from minutes to years. Lastly, the location encoder, analogous to the background rate of seismicity in the ETAS model, learns location-specific information. Details of the encoder architectures are given in the methods section below, and the source code is available at36.

Here we apply the FERN model to the observed seismicity of the greater Japanese Islands region recorded over the last 30 years. The study region is discretized into a grid of square cells of dimension \(0.25^{\circ } \times 0.25^{\circ }\). The input to the model is a catalog of earthquakes, including the hypocenter, magnitude, and time of each event, as well as the geographic location of the gridded cell centers. This information is passed through three neural encoders to generate a latent representation of seismic history. The encoded history is then passed through a neural decoder to perform the prediction task.

The three study regions in northern Japan. Earthquakes larger than \(M_w=5\) that occurred during the study period are plotted. Maps, here and in Fig. 4, were generated using the pygmt37 and matplotlib38 python packages.

We apply the FERN model to study the seismic activity in three sub-regions near the Japan subduction zone (Fig. 2). Using hypocenter data from the JMA earthquake catalog39 the network is trained separately in each region using strict train-validation-test temporal splits of the data with a training period spanning the years 1979-1995 and a validation period of 1996–2003. A hyper-parameter search is performed to determine the optimal network parameters. Finally, the best performing model is trained over both the training and validation period, and is evaluated over the catalog of the years 2004–2011 (test period). The evaluation is performed over a finer grid, \(0.05^{\circ } \times 0.05^{\circ }\), to obtain a better estimation of model performance. Numerical tests have demonstrated that further resolution refinement does not improve our estimation of the log-likelihood. All metrics reported below pertain to the performance of the FERN model during a test period that ends prior to the great Tohoku-oki earthquake of March 2011. Simultaneously, we also train an ETAS model26,40,41,42 over the same temporal and spatial windows. Average seismicity rates in the three period are given for each region in Table I of the supplementary material.

As a first step, we train the FERN model to predict the instantaneous rate of seismicity, \(\lambda (x,y,t \mid H_{t-})\) which is also the output of the ETAS model. The network is trained to optimize the log-likelihood of the observed catalog, \({\mathscr {L}} = \sum _i \log \lambda _i-\iiint \lambda (x,y,t)dx\,dy\,dt\)13,15 where \(\lambda _i = \lambda (x_i, y_i, t_i \mid H_{t_i-})\) is the predicted rate at the spatiotemporal location of the i-th earthquake and the sum is taken over all earthquakes in the study region above a certain magnitude cutoff \(M_c\) which we assume to be the estimated completeness magnitude of the catalog.

We find that in all three study regions FERN exhibits a comparable log-likelihood score to that of ETAS (Table 1). Because FERN enables the incorporation of additional information without modification of the model architecture, we can directly include potentially precursory seismic activity from earthquakes of magnitude lower than \(M_c\) using these smaller events only as features, but not as labels. That is, the low-magnitude earthquakes are included as input to the model, but do not change the calculation of \({\mathscr {L}}\). This allows a proper statistical comparison of the model including smaller events (FERN+) with ETAS and with FERN, as all these models describe the same statistical space, namely seismicity above \(M_c\). The additions of smaller magnitude seismicity improves the information gain per earthquake by 4-12% in all tested regions as compared to both ETAS and FERN with large earthquakes only (Table 1). This amounts to \(\sim 0.1\) information bits per earthquake on average.

As a second test, we train the FERN model to perform a short-term seismic forecast. Using the same encoders that were trained to perform rate prediction, and without updating their weights, we now train a different decoder that performs a short term forecast for the number of earthquakes of magnitude \(>M_c\) that occur in each spatial \(0.5^{\circ } \times 0.5^{\circ }\) cell. Specifically, the features in each training example are the earthquakes that occurred up to time t and the label for each cell is the number of earthquakes that occurred in it in the 24 h after time t. Unlike rate prediction, this is a standard (supervised) regression problem whose metrics are readily interpretable. We follow the same strict train-validation-test split as above for training the decoders (the encoders are not retrained), and benchmark model results against catalogs generated from the trained ETAS model. We follow the standard protocol26 of generating 100,000 catalogs from ETAS for each day, and calculating the average number of earthquakes in each cell. The results are presented in Table 2.

We compare model performance using Receiver Operating Characteristic analysis (ROC) obtained by thresholding the model output and counting the rate of true positive (TPR) rate and false positive rate (FPR) predictions (TPR here means that at least one earthquake occurred within a grid cell during a target time interval). For example, in region C at a FPR of 20% ETAS provides a TPR of 80% while FERN+ shows a TPR of 90%. Similar results are obtained for region B, while in region A all models show similar performance.

This is also true in other statistical tests, as shown in panel (b) of Fig. 3. In it we compare the likelihood score of the observed seismicity in the test period (“L-test”) assuming the number of earthquakes in each cell follows a Poisson distribution, and the likelihood score when comparing only the spatial distribution of earthquakes over the test period (“S-test”), see supplementary material for more information. We note that performing short term prediction with FERN (or FERN+) requires only a single forward pass of the trained network, while an ETAS prediction requires running a large number of simulations to collect catalog statistics26. This means that FERN+ provides more than a 1000-fold improvement in runtime.

ROC curves for different models in region C. Regions A,B show qualitatively similar results.

It should be noted that the performance of all models, both machine-learned and ETAS, varies across different geographical regions and time windows43, as we see here as well. For example, it is seen that the information gain of FERN+ over ETAS in Region A is relatively small. It is difficult, in general, to interpret why the neural model performs well in one region and less so in others, though we believe that in this case the cause is the change in seismicity statistics between the train+validation periods, on which the models were trained and calibrated, and the test period, for which the metrics are reported. Table I of the supplementary material details these statistics. It shows that region A shows much more \(M_w\ge 7\) earthquakes in the test period (0.88 events/year) than in the train+validation period (0.2/year). Such a dramatic change does not occur in region B or C. Such effects might be mitigated by continuous training of the model (“pseudo-prospective testing”) or by training a model on several regions in parallel. However, it is worth noting that even in region A the neural model achieves comparable metrics to that of ETAS.

Unlike ETAS, the parameters of the neural model cannot be trivially interpreted, which is common for neural models44. However we can experiment with FERN model to answer the question: How does the predicted seismicity rate \(\lambda (x,y,t)\) change in response to a single earthquake? The answer that the ETAS model gives to this basic question is, by definition, f. To answer this question with the FERN model, we added an synthetic earthquake to the event catalog, at an arbitrary time and location in Region A (cf. Fig. 2). In Fig. 4 we present the difference between model prediction for \(\lambda (x,y,t)\) 1 h after this synthetic earthquake and its prediction when this earthquake is not present, for both ETAS and FERN.

We find that the response of FERN shows a complex and anisotropic spatial structure, with increased response along the fault trace. We note that the location of the fault line was not included as a feature to the model and that the FERN model learns that the increased seismic activity is neither isotropic nor spatially homogeneous which is, of course, a well known characteristic of seismicity45,46,47,48,49. It is also seen that the output of the location encoder shows similar spatial patterns to the patterns of seismic activity, as was recently shown50. Similarly we find that the temporal dependence of the the rate increase learned by FERN is a power-law, but one that decays slower than the ETAS prediction, depends less strongly on the magnitude, and the magnitude dependence is not homogeneous but rather spatially dependent (Supplementary material).

Looking inside the model. (a,b) The rate difference, as predicted by ETAS and FERN to a single earthquake. We added a synthetic earthquake to the catalog at \(144^{\circ }, 40^{\circ }\) (marked with a yellow star) at time \(t=10.10.2010\) at midnight, and calculated the difference between the rate predicted by the models with and without the synthetic earthquake, 1 h after the event. The plotted region is Region A, and the fault line is shown in red. (c) The activation of one of the latent neurons in the output of the location encoder, for each spatial cell (other neurons show qualitatively similar patterns). It is seen that this patterns correlates well with total number of earthquakes in the cell, shown in (d). We can think about the output of the location encoder as a generalization of the background rate \(\mu\) of the ETAS model, which is shown in (e).

We present a neural architecture for earthquake rate forecasting, adopting the point-process approach but replacing the assumed functional forms of the ETAS model with learned embeddings. Our method shows comparable or superior test metrics (without uncertainty analysis), and the latent representation of seismic history generated by the neural encoders, which were trained to perform rate prediction, can readily be used also for related tasks with small additional effort. This raises hope that such models could be useful in other tasks, such as magnitude prediction or hazard assessment.

Here we describe the main design choices of the FERN model. Full details can be found in the supplementary material.

Recent earthquakes (ETAS-like): This encoder model is a direct generalization the sum term in the definition of ETAS. That is, its output is a sum of a function applied to cataloged data of every earthquake in the (recent) past. The function is constructed in the following way: The catalog provides 5 numbers that describe each earthquake, indexed by i: the time of the event \(t_i\), its epicentral location \(x_i, y_i\), depth \(d_i\) and moment magnitude \(M_i\). We use UTM coordinates for x, y. In addition, the model has access to the spatiotemporal parameters of the cell x, y, t. For each earthquake and cell we calculate a list of k features \(F^1(t,x,y,t_i,x_i,y_i,d_i,M_i)\dots F^k(t,x,y,t_i,x_i,y_i,d_i,M_i)\). These feature functions are inspired by ETAS and constrained by physical considerations. A few examples of feature functions are the magnitude of the earthquake, \(F_1 = e^{M_i}\); the reciprocal of the elapsed time since the earthquake, \(F^2 = 1 / (t - t_j)\); the reciprocal of the distance earthquake’s epicenter, \(F^3= 1 / \sqrt{(x - x_j)^2 + (y - y_j)^2}\), etc. The full list of feature functions is given in table III of the supplementary material. The feature vector \(\left( F^1_i,\dots ,F^k_i\right)\) is then passed through a multi-layer perceptron44 whose output is a latent representation of the earthquake features. This representation is then summed over the past N earthquakes, like the sum that defines \(\lambda\) in the ETAS model. The encoder is clearly invariant to permutations of catalog rows. Simply put, this encoder essentially mimics the structure of them time-dependent part of an ETAS model, only replacing the function f with a neural network, allowing to parameterize a much larger family of functions.

Long range seismicity: The goal of this encoder is to capture long- and short-term seismicity at the point (x, y) at time t. The features for this model are built as follows. For each such point we calculate n(T, d, M), which is the number of earthquakes with magnitude larger than M, that occurred at most T seconds prior to t, at epicentral distance smaller than d from (x, y). For implementation simplicity we use \(L_1\) distance, but this choice has negligible effects on the results. The parameters T, d, M are taken from a predefined list. The values of T and d are logarithmically spaced, allowing to capture very long histories as well as recent activity. This produces a feature vector \((n_1, \dots , n_k)\) per spatial location. Following a weight-sharing strategy similar to that of the recent earthquake encoder, we then use a multi-layer perceptron to parameterize a function g(n, T, d, M) which is applied to all spatial locations. Implementation details are given in the supplementary material. Our experiments showed that using such weight sharing, i.e. learning a single function g, gives significantly better results then learning a more general model that takes the individual \(n_i\) as input.

Location: This encoder is intended to capture local properties for each spatial cell. The model output is a 16-dimensional vector representing the cell’s identity. In Fig. 4 it is seen that the encoding is well correlated with seismicity. The encoder is implemented as a one-hot encoder44 (treating every cell as a different class), followed by a single fully connected layer.

To calculate the loss,

we use the method suggested by Omi. et. al53. The total train period is divided into intervals that begin an end at the times \(\{t_i\}\) where earthquakes occurred. Each training example corresponds to one such interval \([t_i, t_{i+1}]\). For each interval, the catalog of all earthquakes that occurred prior to \(t_i\) is passed to the different encoders. The output of the encoders, the latent representation of \(H_{t-}\), is then passed to a decoder that outputs \(\int _{t_i}^{t_{i+1}}\lambda dt\) for each cell. For this calculation, \(\Delta t_i=t_{i+1}-t_{i}\) is supplied as in input to the decoder (see Fig. 1). The second term in Eq. (1) is then evaluated by summing the model output over all examples, and the first term is obtained through automatic differentiation, which is computationally cheap in neural networks.

The datasets generated and/or analysed during the current study are available in the Japan Meterological Agency (JMA) earthquake catalog, https://www.data.jma.go.jp/svd/eqev/data/bulletin/index_e.html.

van den Ende, M. P. & Ampuero, J.-P. Automated seismic source characterization using deep graph neural networks. Geophys. Res. Lett. 47, e2020GL088690 (2020).

ADS Google Scholar

Zhang, X. et al. A data-driven framework for automated detection of aircraft-generated signals in seismic array data using machine learning. Seismol. Soc. Am. 93, 226–240 (2022).

Google Scholar

Mousavi, S. M., Ellsworth, W. L., Zhu, W., Chuang, L. Y. & Beroza, G. C. Earthquake transformer—An attentive deep-learning model for simultaneous earthquake detection and phase picking. Nat. Commun. 11, 1–12 (2020).

Article Google Scholar

Ross, Z. E., Trugman, D. T., Hauksson, E. & Shearer, P. M. Searching for hidden earthquakes in southern California. Science 364, 767–771 (2019).

Article ADS CAS PubMed Google Scholar

Bergen, K. J., Johnson, P. A., de Hoop, M. V. & Beroza, G. C. Machine learning for data-driven discovery in solid earth geoscience. Science 363, eaau0323 (2019).

Article PubMed Google Scholar

Kong, Q. et al. Machine learning in seismology: Turning data into insights. Seismol. Res. Lett. 90, 3–14 (2019).

Article Google Scholar

Licciardi, A., Bletery, Q., Rouet-Leduc, B., Ampuero, J.-P. & Juhel, K. Instantaneous tracking of earthquake growth with elastogravity signals. Nature 606, 1–6 (2022).

Article Google Scholar

Mignan, A. & Broccardo, M. Neural network applications in earthquake prediction (1994–2019): Meta-analytic and statistical insights on their limitations. Seismol. Res. Lett. 91, 2330–2342 (2020).

Article Google Scholar

Mancini, S. et al. On the use of high-resolution and deep-learning seismic catalogs for short-term earthquake forecasts: Potential benefits and current limitations. J. Geophys. Res. Solid Earth 127(11), e2022JB025202 (2022).

Article ADS CAS PubMed PubMed Central Google Scholar

Berryman, K. R. et al. Major earthquakes occur regularly on an isolated plate boundary fault. Science 336, 1690–1693 (2012).

Article ADS CAS PubMed Google Scholar

Uchida, N. & Bürgmann, R. Repeating earthquakes. Annu. Rev. Earth Planet. Sci. 47, 305–332 (2019).

Article ADS CAS Google Scholar

Geller, R. J. Earthquake prediction: A critical review. Geophys. J. Int. 131, 425–450 (1997).

Article ADS Google Scholar

Vere-Jones, D. Earthquake prediction—A statistician’s view. J. Phys. Earth 26, 129–146 (1978).

Article Google Scholar

Ogata, Y. Seismicity analysis through point-process modeling: A review. In Seismicity Patterns, Their Statistical Significance and Physical Meaning 471–507 (1999).

Rasmussen, J. G. Lecture notes: Temporal point processes and the conditional intensity function. (2018) arXiv preprint arXiv:1806.00221.

Mignan, A. Seismicity precursors to large earthquakes unified in a stress accumulation framework. Geophys. Res. Lett. 39, L21308 (2012).

Article ADS Google Scholar

Trugman, D. T. & Ross, Z. E. Pervasive foreshock activity across southern California. Geophys. Res. Lett. 46(15), 8772–8781 (2019).

Article ADS Google Scholar

Utsu, T. et al. The centenary of the omori formula for a decay law of aftershock activity. J. Phys. Earth 43, 1–33 (1995).

Article Google Scholar

Field, E. H. et al. A spatiotemporal clustering model for the third uniform California earthquake rupture forecast (UCERF3-ETAS): Toward an operational earthquake forecast. Bull. Seismol. Soc. Am. 107, 1049–1081 (2017).

Article Google Scholar

Ogata, Y. Statistical models for earthquake occurrences and residual analysis for point processes. J. Am. Stat. Assoc. 83, 9–27 (1988).

Article Google Scholar

Kumazawa, T. & Ogata, Y. Nonstationary ETAS models for nonstandard earthquakes. Ann. Appl. Stat. 8, 1825–1852 (2014).

Article MathSciNet MATH Google Scholar

Ogata, Y. & Zhuang, J. Space-time ETAS models and an improved extension. Tectonophysics 413, 13–23 (2006).

Article ADS Google Scholar

Segou, M., Parsons, T. & Ellsworth, W. Comparative evaluation of physics-based and statistical forecasts in northern California. J. Geophys. Res. Solid Earth 118, 6219–6240 (2013).

Article ADS Google Scholar

Kovchegov, Y., Zaliapin, I. & Ben-Zion, Y. Invariant Galton-Watson branching process for earthquake occurrence. Geophys. J. Int. 231, 567–583 (2022).

Article ADS Google Scholar

Werner, M. J., Helmstetter, A., Jackson, D. D. & Kagan, Y. Y. High-resolution long-term and short-term earthquake forecasts for California. Bull. Seismol. Soc. Am. 101, 1630–1648 (2011).

Article Google Scholar

Zhuang, J. Next-day earthquake forecasts for the japan region generated by the etas model. Earth Planets Space 63, 207–216 (2011).

Article ADS Google Scholar

Llenos, A. L. & Michael, A. J. Ensembles of etas models provide optimal operational earthquake forecasting during swarms: Insights from the 2015 san ramon, california swarmensembles of etas models provide optimal operational earthquake forecasting during swarms. Bull. Seismol. Soc. Am. 109, 2145–2158 (2019).

Article Google Scholar

Milner, K. R., Field, E. H., Savran, W. H., Page, M. T. & Jordan, T. H. Operational earthquake forecasting during the 2019 ridgecrest, California, earthquake sequence with the ucerf3-etas model. Seismol. Res. Lett. 91, 1567–1578 (2020).

Article Google Scholar

Veen, A. & Schoenberg, F. P. Estimation of space-time branching process models in seismology using an EM-type algorithm. J. Am. Stat. Assoc. 103, 614–624 (2008).

Article MathSciNet CAS MATH Google Scholar

Zhuang, J., Ogata, Y. & Wang, T. Data completeness of the Kumamoto earthquake sequence in the JMA catalog and its influence on the estimation of the ETAS parameters. Earth Planets Space 69, 1–12 (2017).

Article ADS Google Scholar

Seif, S., Mignan, A., Zechar, J. D., Werner, M. J. & Wiemer, S. Estimating ETAS: The effects of truncation, missing data, and model assumptions. J. Geophys. Res. Solid Earth 122, 449–469 (2017).

Article ADS Google Scholar

Schoenberg, F. P., Chu, A. & Veen, A. On the relationship between lower magnitude thresholds and bias in epidemic-type aftershock sequence parameter estimates. J. Geophys. Res. 115(B4), B04309 (2010).

ADS Google Scholar

Harte, D. S. Model parameter estimation bias induced by earthquake magnitude cut-off. Geophys. J. Int. 204, 1266–1287 (2016).

Article ADS Google Scholar

Mizrahi, L., Nandan, S. & Wiemer, S. Embracing data incompleteness for better earthquake forecasting. J. Geophys. Res. Solid Earth 126, e2021JB022379 (2021).

Article ADS Google Scholar

Adelfio, G. & Chiodi, M. Including covariates in a space-time point process with application to seismicity. Stat. Methods Appl. 30, 947–971 (2020).

Article MathSciNet MATH Google Scholar

Zlydenko, O. et al. https://github.com/google-research/google-research/tree/master/earthquakes_fern.

Uieda, L. et al. PyGMT: A Python interface for the Generic Mapping Tools (2023). https://doi.org/10.5281/zenodo.7772533.

Hunter, J. D. Matplotlib: A 2D graphics environment. Comput. Science Eng. 9, 90–95 (2007).

Article Google Scholar

Japan Meteorological Agency. The Seismological Bulletin of Japan. https://www.data.jma.go.jp/svd/eqev/data/bulletin/index_e.html.

Zhuang, J., Ogata, Y. & Vere-Jones, D. Stochastic declustering of space-time earthquake occurrences. J. Am. Stat. Assoc. 97, 369–380 (2002).

Article MathSciNet MATH Google Scholar

Zhuang, J., Ogata, Y. & Vere-Jones, D. Analyzing earthquake clustering features by using stochastic reconstruction. J. Geophys. Res. Solid Earth 109(B5), B05301 (2004).

Article ADS Google Scholar

Zhuang, J. Second-order residual analysis of spatiotemporal point processes and applications in model evaluation. J. R. Stat. Soc. Ser. B Stat. Methodol. 68, 635–653 (2006).

Article MathSciNet MATH Google Scholar

Bayona, J. A. et al. Are regionally calibrated seismicity models more informative than global models? Insights from California, new zealand, and italy. Seismic Rec. 3, 86–95 (2023).

Article Google Scholar

Murphy, K. P. Machine Learning: A Probabilistic Perspective (MIT Press, 2012).

MATH Google Scholar

Parsons, T., Stein, R. S., Simpson, R. W. & Reasenberg, P. A. Stress sensitivity of fault seismicity: A comparison between limited-offset oblique and major strike-slip faults. J. Geophys. Res. Solid Earth 104, 20183–20202 (1999).

Article Google Scholar

Toda, S., Stein, R. S., Reasenberg, P. A., Dieterich, J. H. & Yoshida, A. Stress transferred by the 1995 mw = 6.9 Kobe, Japan, shock: Effect on aftershocks and future earthquake probabilities. J. Geophys. Res. Solid Earth 103, 24543–24565 (1998).

Article Google Scholar

King, G. C., Stein, R. S. & Lin, J. Static stress changes and the triggering of earthquakes. Bull. Seismol. Soc. Am. 84, 935–953 (1994).

Google Scholar

Yabe, S. & Ide, S. Why do aftershocks occur within the rupture area of a large earthquake?. Geophys. Res. Lett. 45, 4780–4787 (2018).

Article ADS Google Scholar

Ross, Z. E., Ben-Zion, Y. & Zaliapin, I. Geometrical properties of seismicity in California. Geophys. J. Int. 231, 493–504 (2022).

Article ADS Google Scholar

Page, M. T. & van der Elst, N. J. Aftershocks preferentially occur in previously active areas. Seismic Rec. 2, 100–106 (2022).

Article Google Scholar

Zechar, J. D. et al. The collaboratory for the study of earthquake predictability perspective on computational earthquake science. Concurr. Comput. Pract. Exp. 22, 1836–1847 (2010).

Article Google Scholar

Jordan, T. H. Earthquake predictability, brick by brick. Seismol. Res. Lett. 77, 3–6 (2006).

Article Google Scholar

Omi, T. et al. Implementation of a real-time system for automatic aftershock forecasting in japan. Seismol. Res. Lett. 90, 242–250 (2019).

Article Google Scholar

Download references

This study was funded by Israel Science Foundation (Grant No. 1907/22).

Google Research, Tel-Aviv, Israel

Oleg Zlydenko, Gal Elidan, Avinatan Hassidim, Doron Kukliansky, Yossi Matias, Alexandra Molchanov, Sella Nevo & Yohai Bar-Sinai

Google Research, Cambridge, MA, USA

Brendan Meade

Department of Earth and Planetary Sciences, Harvard University, Cambridge, MA, USA

Brendan Meade

Department of Condensed Matter Physics, Tel-Aviv University, Tel-Aviv, Israel

Yohai Bar-Sinai

You can also search for this author in PubMed Google Scholar

Y.B.S., O.Z., S.N., B.M., G.E., A.H. and Y.M. designed research. O.Z., Y.B.S., D.K., S.N. and A.M. performed research. O.Z., B.M. and Y.B.S. wrote the manuscript. All authors reviewed the manuscript.

Correspondence to Yohai Bar-Sinai.

The authors declare no competing interests.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and Permissions

Zlydenko, O., Elidan, G., Hassidim, A. et al. A neural encoder for earthquake rate forecasting. Sci Rep 13, 12350 (2023). https://doi.org/10.1038/s41598-023-38033-9

Download citation

Received: 17 January 2023

Accepted: 01 July 2023

Published: 31 July 2023

DOI: https://doi.org/10.1038/s41598-023-38033-9

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

Wavelet gated multiformer for groundwater time series forecasting

Sinusoid Pty Ltd Introduces RPE4 Absolute Analog 4

News

A neural encoder for earthquake rate forecasting