arXiv subjects

Andrea Donnellan

Publications and source records attributed to Andrea Donnellan.

9 recordsLinked to original sources

arXiv2026

Exceedance Probabilities for Large Earthquakes From DIY Local Earthquake Ensemble Nowcasting and Forecasting

This paper focuses on the problem of anticipating the local occurrence of future large earthquakes. "Local" is defined as the probability of a large earthquake occurring with a defined circle of arbitrary radius surrounding a point of interest. The main (and for that matter, the only) assumption for all these works is that the Gutenberg-Richter (GR) magnitude-frequency relation holds. Here we describe a method for computing calendar time forecasts in a local area for large earthquakes of a target magnitude MT using a count small earthquakes MS < MT in the area. Using the idea that the GR relation is valid throughout the surrounding region, we define an ensemble of earthquakes in larger surrounding regions to be used in computing the forecast. What follows is simple data mining. The method has significant skill, as defined by the Receiver Operating Characteristic (ROC) test, which improves as time since the last major earthquake increases. The probability is conditioned on the number of small earthquakes n(t) that have occurred since the last large earthquake. The probability is computed directly as the Positive Predictive Value (PPV) associated with the ROC curve. The method is validated by comparison to the UCERF3 forecasts for the UCERF3-defined geographic boxes centered on Los Angeles and San Francisco. The method is then applied to a 125-KM radius circular area around Los Angeles, California, following the January 17, 1994 magnitude M6.7 Northridge earthquake, and short term forecasts (1 year and 5 year ) are computed.

physics.geo-ph↗

arXiv2025

Calendar Time Local Earthquake Forecasts from Earthquake Nowcasts: A Do-It-Yourself (DIY) Ensemble Method

This paper presents a new technical method for computing calendar time forecasts in a local area for large earthquakes of a target magnitude MT using a count small earthquakes MS < MT in the area, together with the Gutenberg-Richter (GR) magnitude-frequency relation. The GR relation states that for every large target earthquake of magnitude greater than MT , there are on average NGR small earthquakes of magnitude MT > M >= MS. The only assumption is that the GR statistics of the local area are the same as in the larger surrounding regions. This assumption is used to construct an ensemble of earthquakes in larger surrounding regions to be used in computing the forecast. The method has significant skill, as defined by the Receiver Operating Characteristic (ROC) test, which improves as time since the last major earthquake increases. The probability is conditioned on the number of small earthquakes n(t) that have occurred since the last large earthquake. There is no need to assume a probability model, as the probability is instead computed directly as the Positive Predictive Value (PPV) associated with the ROC curve. The method is validated by comparison to the UCERF3 forecasts for the UCERF3-defined geographic boxes centered on Los Angeles and San Francisco. The method is then applied to a 125-KM radius circular area around Los Angeles, California, following the January 17, 1994 magnitude M6.7 Northridge earthquake, and short term forecasts (1 year and 5 year ) are computed.

physics.geo-ph↗

arXiv2025

From Local Earthquake Nowcasting to Natural Time Forecasting: A Simple Do-It-Yourself (DIY) Method

Previous papers have outlined nowcasting methods to track the current state of earthquake hazard using only observed seismic catalogs. The basis for one of these methods, the "counting method", is the Gutenberg-Richter (GR) magnitude-frequency relation. The GR relation states that for every large earthquake of magnitude greater than MT , there are on average NGR small earthquakes of magnitude MS. In this paper we use this basic relation, combined with the Receiver Operating Characteristic (ROC) formalism from machine learning, to compute the probability of a large earthquake. The probability is conditioned on the number of small earthquakes n(t) that have occurred since the last large earthquake. We work in natural time, which is defined as the count of small earthquakes between large earthquakes. We do not need to assume a probability model, which is a major advantage. Instead, the probability is computed as the Positive Predictive Value (PPV) associated with the ROC curve. We find that the PPV following the last large earthquake initially decreases as more small earthquakes occur, indicating the property of temporal clustering of large earthquakes as is observed. As the number of small earthquakes continues to accumulate, the PPV subsequently begins to increase. Eventually a point is reached beyond which the rate of increase becomes much larger and more dramatic. Here we describe and illustrate the method by applying it to a local region around Los Angeles, California, following the January 17, 1994 magnitude M6.7 Northridge earthquake.

physics.geo-ph↗

arXiv2024

Time Series Foundation Models and Deep Learning Architectures for Earthquake Temporal and Spatial Nowcasting

Advancing the capabilities of earthquake nowcasting, the real-time forecasting of seismic activities remains a crucial and enduring objective aimed at reducing casualties. This multifaceted challenge has recently gained attention within the deep learning domain, facilitated by the availability of extensive, long-term earthquake datasets. Despite significant advancements, existing literature on earthquake nowcasting lacks comprehensive evaluations of pre-trained foundation models and modern deep learning architectures. These architectures, such as transformers or graph neural networks, uniquely focus on different aspects of data, including spatial relationships, temporal patterns, and multi-scale dependencies. This paper addresses the mentioned gap by analyzing different architectures and introducing two innovation approaches called MultiFoundationQuake and GNNCoder. We formulate earthquake nowcasting as a time series forecasting problem for the next 14 days within 0.1-degree spatial bins in Southern California, spanning from 1986 to 2024. Earthquake time series is forecasted as a function of logarithm energy released by quakes. Our comprehensive evaluation employs several key performance metrics, notably Nash-Sutcliffe Efficiency and Mean Squared Error, over time in each spatial region. The results demonstrate that our introduced models outperform other custom architectures by effectively capturing temporal-spatial relationships inherent in seismic data. The performance of existing foundation models varies significantly based on the pre-training datasets, emphasizing the need for careful dataset selection. However, we introduce a new general approach termed MultiFoundationPattern that combines a bespoke pattern with foundation model results handled as auxiliary streams. In the earthquake case, the resultant MultiFoundationQuake model achieves the best overall performance.

cs.LG↗

arXiv2024

Nowcasting Earthquakes with QuakeGPT: Methods and First Results

Earthquake nowcasting has been proposed as a means of tracking the change in large earthquake potential in a seismically active area. The method was developed using observable seismic data, in which probabilities of future large earthquakes can be computed using Receiver Operating Characteristic (ROC) methods. Furthermore, analysis of the Shannon information content of the earthquake catalogs has been used to show that there is information contained in the catalogs, and that it can vary in time. Here we discuss a new method for earthquake nowcasting that uses an AI-enhanced deep learning model "QuakeGPT" that is based on an attention-based science transformer adapted for time series forecasting. Such dot product attention-based transformers were introduced by Vaswani et al. (2017), and are the basis for the new large language models such as ChatGPT. To use these science transformers, they must first be trained on a large corpus of data. A problem is that the existing history of reliable earthquake catalog data extends back in time only a few decades, which is almost certainly too short to train a model for reliable nowcasting/forecasting. As a result, we turn to earthquake simulations to train the transformer model. Specifically we discuss a simple stochastic earthquake simulation model "ERAS" that has recently been introduced. The ERAS model is similar to the more common "ETAS" models, the difference being that the ERAS model has only 2 novel, adjustable parameters, rather than the 6-8 adjustable parameters that characterize most ETAS models. Using this ERAS model, we then define a transformer model and train it using a long catalog of ERAS simulations, then apply it to an ERAS validation dataset with the transformer model. In this paper, we describe this new method and assess the applicability to observed earthquake catalogs for use in nowcasting/forecasting.

physics.geo-ph↗

arXiv2022

Does the Catalog of California Earthquakes, with Aftershocks Included, Contain Information about Future Large Earthquakes?

Yes. Interval statistics have been used to conclude that major earthquakes are random events in time and cannot be anticipated or predicted. Machine learning is a powerful new technique that enhances our ability to understand the information content of earthquake catalogs. We show that catalogs contain significant information on current hazard and future predictability for large earthquakes.

physics.geo-ph↗

arXiv2021

Earthquake Nowcasting with Deep Learning

We review previous approaches to nowcasting earthquakes and introduce new approaches based on deep learning using three distinct models based on recurrent neural networks and transformers. We discuss different choices for observables and measures presenting promising initial results for a region of Southern California from 1950-2020. Earthquake activity is predicted as a function of 0.1-degree spatial bins for time periods varying from two weeks to four years. The overall quality is measured by the Nash Sutcliffe Efficiency comparing the deviation of nowcast and observation with the variance over time in each spatial region. The software is available as open-source together with the preprocessed data from the USGS.

physics.geo-ph↗

arXiv2005

Systematic procedural and sensitivity analysis of the pattern informatics method for forecasting large (M > 5) earthquake events in southern California

Recent studies in the literature have introduced a new approach to earthquake forecasting based on representing the space-time patterns of localized seismicity by a time-dependent system state vector in a real-valued Hilbert space and deducing information about future space-time fluctuations from the phase angle of the state vector. While the success rate of this Pattern Informatics (PI) method has been encouraging, the method is still in its infancy. Procedural analysis, statistical testing, parameter sensitivity investigation and optimization all still need to be performed. In this paper, we attempt to optimize the PI approach by developing quantitative values for "predictive goodness" and analyzing possible variations in the proposed procedure. In addition, we attempt to quantify the systematic dependence on the quality of the input catalog of historic data and develop methods for combining catalogs from regions of different seismic rates.

physics.geo-ph↗

arXiv2005

Modification of the pattern informatics method for forecasting large earthquake events using complex eigenvectors

Recent studies have shown that real-valued principal component analysis can be applied to earthquake fault systems for forecasting and prediction. In addition, theoretical analysis indicates that earthquake stresses may obey a wave-like equation, having solutions with inverse frequencies for a given fault similar to those that characterize the time intervals between the largest events on the fault. It is therefore desirable to apply complex principal component analysis to develop earthquake forecast algorithms. In this paper we modify the Pattern Informatics method of earthquake forecasting to take advantage of the wave-like properties of seismic stresses and utilize the Hilbert transform to create complex eigenvectors out of measured time series. We show that Pattern Informatics analyses using complex eigenvectors create short-term forecast hot-spot maps that differ from hot-spot maps created using only real-valued data and suggest methods of analyzing the differences and calculating the information gain.

physics.geo-ph↗