Gustavo Bejar
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Gustavo Bejar has a total of 13 co-authors. Here are the top co-authors by number of shared publications:
- Gregory P. Waite
Department of Geological and Mining Engineering and Sciences Michigan Technological University Houghton MI USA
5 shared publications - Jeffrey B. Johnson
Boise State University
4 shared publications - Amilcar Roca
INSIVUMEH
3 shared publications - Ashley R. Bosa
Boise State University
3 shared publications - Jerry Mock
Boise State University
3 shared publications Armando Pineda
Independent
2 shared publications
Latest Publications
- 11 February 2026
- Published by American Geophysical Union (AGU) in Journal of Geophysical Research: Solid Earth
- Vol. 131 (2)
Seismic Characterization of Lahars on Volcán de Fuego Toward the Development of a Machine Learning‐Based Detection Algorithm
Lahars are among the most frequent hazards associated with Volcán de Fuego, Guatemala. Despite their recurrence, early detection and automated alerts remain challenging since they often rely on manual monitoring and sparse visual confirmation. Yet, we can harness the high number of flows triggered every rainy season to characterize their seismic signatures and quantify their size and behavior. For this, we used seismic stations located along two active lahar channels on Fuego where this characterization describes a somewhat stable long‐term flow behavior. This work revealed more varied short‐term behavior characterized by increasing seismic activity in the time domain and a shift toward lower frequencies as these flows propagate downstream. Building on this characterization, we implemented K‐nearest neighbor (KNN) based detectors using seismic signal attributes describing samples of the data in the time and frequency domains, as well as statistical functions of these samples. We trained generalized and station‐specific detectors that achieved high accuracy for detecting moderate‐to‐large flows with lower performance for smaller or ambiguous events. We found that root mean square amplitude, a proxy for flow size, appears to control detector performance more than other signal features. The detector is computationally efficient and, in the case of Fuego, did not require additional instrumentation. This framework presents a portable solution for enhancing automated lahar detection while minimizing the use of location‐specific parameters required by other methods.
- 11 February 2026
- Published by American Geophysical Union (AGU) in Journal of Geophysical Research: Solid Earth
- Vol. 131 (2)
Seismic Characterization of Lahars on Volcán de Fuego Toward the Development of a Machine Learning‐Based Detection Algorithm
Lahars are among the most frequent hazards associated with Volcán de Fuego, Guatemala. Despite their recurrence, early detection and automated alerts remain challenging since they often rely on manual monitoring and sparse visual confirmation. Yet, we can harness the high number of flows triggered every rainy season to characterize their seismic signatures and quantify their size and behavior. For this, we used seismic stations located along two active lahar channels on Fuego where this characterization describes a somewhat stable long‐term flow behavior. This work revealed more varied short‐term behavior characterized by increasing seismic activity in the time domain and a shift toward lower frequencies as these flows propagate downstream. Building on this characterization, we implemented K‐nearest neighbor (KNN) based detectors using seismic signal attributes describing samples of the data in the time and frequency domains, as well as statistical functions of these samples. We trained generalized and station‐specific detectors that achieved high accuracy for detecting moderate‐to‐large flows with lower performance for smaller or ambiguous events. We found that root mean square amplitude, a proxy for flow size, appears to control detector performance more than other signal features. The detector is computationally efficient and, in the case of Fuego, did not require additional instrumentation. This framework presents a portable solution for enhancing automated lahar detection while minimizing the use of location‐specific parameters required by other methods.
- Open Access
- 27 May 2025
- Published by Wiley
Seismic Characterization of Lahars on Volcán de Fuego Toward the Development of a Machine Learning-Based Detection Algorithm
Lahars are among the most frequent hazards associated with Volcán de Fuego, Guatemala. Despite their recurrence, early detection and automated alerts remain challenging since they often rely on manual monitoring and sparse visual confirmation. Yet, we can harness the high number of flows triggered every rainy season to characterize their seismic signatures and quantify their size and behavior. For this, we used seismic stations located along two active lahar channels on Fuego where this characterization describes a somewhat stable long-term flow behavior. This work revealed more varied short term behavior characterized by increasing seismic activity in the time domain and a shift toward lower frequencies as these flows propagate downstream. Building on this characterization, we implemented K-nearest neighbor (KNN)-based detectors using seismic signal attributes describing samples of the data in the time and frequency domains, as well as statistical functions of these samples. We trained generalized and station-specific detectors that achieved high accuracy for detecting moderate-to-large flows with lower performance for smaller or ambiguous events. We found that root mean square amplitude, a proxy for flow size, appears to control detector performance more than other signal features. The detector is computationally efficient and, in the case of Fuego, did not require additional instrumentation. This framework presents a portable solution for enhancing automated lahar detection while minimizing the use of location-specific parameters required by other methods.
- 27 May 2025
- Published by Wiley
Open AccessSeismic Characterization of Lahars on Volcán de Fuego Toward the Development of a Machine Learning-Based Detection Algorithm
Lahars are among the most frequent hazards associated with Volcán de Fuego, Guatemala. Despite their recurrence, early detection and automated alerts remain challenging since they often rely on manual monitoring and sparse visual confirmation. Yet, we can harness the high number of flows triggered every rainy season to characterize their seismic signatures and quantify their size and behavior. For this, we used seismic stations located along two active lahar channels on Fuego where this characterization describes a somewhat stable long-term flow behavior. This work revealed more varied short term behavior characterized by increasing seismic activity in the time domain and a shift toward lower frequencies as these flows propagate downstream. Building on this characterization, we implemented K-nearest neighbor (KNN)-based detectors using seismic signal attributes describing samples of the data in the time and frequency domains, as well as statistical functions of these samples. We trained generalized and station-specific detectors that achieved high accuracy for detecting moderate-to-large flows with lower performance for smaller or ambiguous events. We found that root mean square amplitude, a proxy for flow size, appears to control detector performance more than other signal features. The detector is computationally efficient and, in the case of Fuego, did not require additional instrumentation. This framework presents a portable solution for enhancing automated lahar detection while minimizing the use of location-specific parameters required by other methods.
- Open Access
- 28 September 2024
- Published by Springer Nature in Natural Hazards
- Vol. 121 (3) , 3431-3472
Dynamics of rain-triggered lahars and destructive power inferred from seismo-acoustic arrays and time-lapse camera correlation at Volcán de Fuego, Guatemala
Lahars, or volcanic mudflows, are one of the most devastating natural, volcanic hazards. Deadly lahars, such as the one that occurred after the Nevado del Ruiz, Columbia eruption in 1985, in which at least 23,000 people tragically lost their lives, threaten the safety and well-being of humans, the economy, and the infrastructure of many of the communities living in the vicinity of volcanoes. Due to their complex flow behaviors, lahars remain a major challenge to those studying them. We present an analysis of several rain-triggered lahar events at Volcán Fuego in Guatemala using both seismic and infrasound monitoring to quantify both ground vibrations and low-frequency atmospheric sound waves associated with these mudflows. Geophysical data collected over this field campaign quantifies flow parameters such as velocities, stage and the frequency of these rain-triggered lahars. Time-lapse imagery of lahar flows is compared with filtered seismo-acoustic signal characteristics to ascertain stage predictions and relationship to stage fluxes. Using random forest regression models, we establish moderate correlations (correlation coefficient modes 0.48–0.53) with statistical significance (p value = 0.01–0.02) between signal energetics and respective stage. Compiling a catalog of rain-triggered lahar events in Volcán de Fuego’s drainages over a season permits a dataset amenable to statistical analysis. Our goal is the development of new-generation geophysical monitoring tools that will be capable of remote and real-time estimation of flow parameters.
- 28 September 2024
- Published by Springer Nature in Natural Hazards
- Vol. 121 (3) , 3431-3472
Open AccessDynamics of rain-triggered lahars and destructive power inferred from seismo-acoustic arrays and time-lapse camera correlation at Volcán de Fuego, Guatemala
Lahars, or volcanic mudflows, are one of the most devastating natural, volcanic hazards. Deadly lahars, such as the one that occurred after the Nevado del Ruiz, Columbia eruption in 1985, in which at least 23,000 people tragically lost their lives, threaten the safety and well-being of humans, the economy, and the infrastructure of many of the communities living in the vicinity of volcanoes. Due to their complex flow behaviors, lahars remain a major challenge to those studying them. We present an analysis of several rain-triggered lahar events at Volcán Fuego in Guatemala using both seismic and infrasound monitoring to quantify both ground vibrations and low-frequency atmospheric sound waves associated with these mudflows. Geophysical data collected over this field campaign quantifies flow parameters such as velocities, stage and the frequency of these rain-triggered lahars. Time-lapse imagery of lahar flows is compared with filtered seismo-acoustic signal characteristics to ascertain stage predictions and relationship to stage fluxes. Using random forest regression models, we establish moderate correlations (correlation coefficient modes 0.48–0.53) with statistical significance (p value = 0.01–0.02) between signal energetics and respective stage. Compiling a catalog of rain-triggered lahar events in Volcán de Fuego’s drainages over a season permits a dataset amenable to statistical analysis. Our goal is the development of new-generation geophysical monitoring tools that will be capable of remote and real-time estimation of flow parameters.
- 19 December 2022
- Published by Boise State University, Albertsons Library
Dataset for Infrasound Early Warning Detections for Lahars
- 19 December 2022
- Published by Boise State University, Albertsons Library
Dataset for Infrasound Early Warning Detections for Lahars