Jeffrey B. Johnson
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Research Interests
Jeffrey B. Johnson has a total of 117 co-authors. Here are the top co-authors by number of shared publications:
- Jacob Fortner Anderson
Department of Geosciences Boise State University Boise ID USA
16 shared publications - Leighton M. Watson
University of Canterbury Christchurch New Zealand
9 shared publications - Gregory P. Waite
Department of Geological and Mining Engineering and Sciences Michigan Technological University Houghton MI USA
5 shared publications - Amilcar Roca
INSIVUMEH
4 shared publications - Bryan Blake Rosenblatt
Boise State University
4 shared publications - Gustavo Bejar
Michigan Technological University
4 shared publications
Latest Publications
- 18 November 2026
- Published by Boise State University, Albertsons Library
Dataset Associated with Technical Paper
- 18 November 2026
- Published by Boise State University, Albertsons Library
Dataset Associated with Technical Paper
- 24 July 2026
- Published by Boise State University, Albertsons Library
Dataset for 'Variability of Howitzer Artillery Infrasound Recorded in Little Cottonwood Canyon, Utah'
These data are a supplement to the MS thesis, “Variability of Howitzer Artillery Infrasound Recorded in Little Cottonwood Canyon, Utah”, by Department of Geosciences student, Owen Walsh. Data provided are for howitzer artillery infrasound signals recorded by six infrasound microphones (infraBSU v2) from March 22 through April 5, 2023 in Little Cottonwood Canyon, Utah, USA. This dataset consists of 142, 20-second signals sampled at 100 Hz. Data provided have been filtered using a 0.5 Hz (2-second) highpass, acausal, 2-pole butterworth filter.
- 24 July 2026
- Published by Boise State University, Albertsons Library
Dataset for 'Variability of Howitzer Artillery Infrasound Recorded in Little Cottonwood Canyon, Utah'
These data are a supplement to the MS thesis, “Variability of Howitzer Artillery Infrasound Recorded in Little Cottonwood Canyon, Utah”, by Department of Geosciences student, Owen Walsh. Data provided are for howitzer artillery infrasound signals recorded by six infrasound microphones (infraBSU v2) from March 22 through April 5, 2023 in Little Cottonwood Canyon, Utah, USA. This dataset consists of 142, 20-second signals sampled at 100 Hz. Data provided have been filtered using a 0.5 Hz (2-second) highpass, acausal, 2-pole butterworth filter.
- Open Access
- 26 June 2026
- Published by American Geophysical Union (AGU) in Geophysical Research Letters
- Vol. 53 (13)
Tracking a Rangeland Fire With Infrasound Arrays
We present the first infrasound analyses of a rangeland fire and demonstrate infrasound's potential in monitoring wildland fires. We recorded a prescribed fire in sagebrush steppe in cool, moist conditions, using eight infrasound arrays of diverse size (3–44 sensors, 10–200 m width) and distance (0–20 km), including four arrays inside the burn area. Infrasound arrays within 2 km of fires detected coherent infrasound between 1 and 10 Hz, both during active ignitions and also while existing fires continued burning during breaks in ignitions. During periods of active ignitions, the direction of arrival estimated by a 44‐sensor array closely tracked the actual locations where the helicopter was igniting vegetation. Despite being farther from the fire than four smaller arrays, the 44‐sensor array performed best at identifying coherent infrasound and yielding consistent directions of arrival. Our results show infrasound's potential for tracking low‐intensity wildland fires in research or hazard monitoring.
- 26 June 2026
- Published by American Geophysical Union (AGU) in Geophysical Research Letters
- Vol. 53 (13)
Open AccessTracking a Rangeland Fire With Infrasound Arrays
We present the first infrasound analyses of a rangeland fire and demonstrate infrasound's potential in monitoring wildland fires. We recorded a prescribed fire in sagebrush steppe in cool, moist conditions, using eight infrasound arrays of diverse size (3–44 sensors, 10–200 m width) and distance (0–20 km), including four arrays inside the burn area. Infrasound arrays within 2 km of fires detected coherent infrasound between 1 and 10 Hz, both during active ignitions and also while existing fires continued burning during breaks in ignitions. During periods of active ignitions, the direction of arrival estimated by a 44‐sensor array closely tracked the actual locations where the helicopter was igniting vegetation. Despite being farther from the fire than four smaller arrays, the 44‐sensor array performed best at identifying coherent infrasound and yielding consistent directions of arrival. Our results show infrasound's potential for tracking low‐intensity wildland fires in research or hazard monitoring.
- Open Access
- 1 March 2026
- Published by Acoustical Society of America (ASA) in The Journal of the Acoustical Society of America
- Vol. 159 (3) , 2795-2809
A deep learning approach for infrasound signal detection in snow-covered terrain
Infrasound signal detection using arrays is an established method for detecting signals from a variety of cultural and geophysical sources. Array processing requires analysis from two or more sensors at an infrasound station. By employing array techniques, signals are isolated from various sources of noise by computing coherence between recorded signals across spatially distributed sensors at a station. Array processing requires multiple sensors and intra-array comparisons that potentially increases time required for event detection. This study presents a signal detection method utilizing deep learning to identify high-quality signals in infrasound waveforms with data from a single infrasound sensor, which might permit event discrimination without using arrays. The approach involves using audio signal processing techniques, applied to infrasound frequencies, to extract a feature space, which is then used to train a deep learning model. A 25-day continuous dataset featuring diverse infrasonic events that include potential avalanches, vehicle traffic, and explosions, collected from Little Cottonwood Canyon, Utah, USA was analyzed for this study. The data provide a convenient testbed for high-quality signal identification using single channel infrasound. The sensor network was motivated by the goal of identifying snow avalanches. Performance was assessed for an optimized model over a range of test data.
- 1 March 2026
- Published by Acoustical Society of America (ASA) in The Journal of the Acoustical Society of America
- Vol. 159 (3) , 2795-2809
Open AccessA deep learning approach for infrasound signal detection in snow-covered terrain
Infrasound signal detection using arrays is an established method for detecting signals from a variety of cultural and geophysical sources. Array processing requires analysis from two or more sensors at an infrasound station. By employing array techniques, signals are isolated from various sources of noise by computing coherence between recorded signals across spatially distributed sensors at a station. Array processing requires multiple sensors and intra-array comparisons that potentially increases time required for event detection. This study presents a signal detection method utilizing deep learning to identify high-quality signals in infrasound waveforms with data from a single infrasound sensor, which might permit event discrimination without using arrays. The approach involves using audio signal processing techniques, applied to infrasound frequencies, to extract a feature space, which is then used to train a deep learning model. A 25-day continuous dataset featuring diverse infrasonic events that include potential avalanches, vehicle traffic, and explosions, collected from Little Cottonwood Canyon, Utah, USA was analyzed for this study. The data provide a convenient testbed for high-quality signal identification using single channel infrasound. The sensor network was motivated by the goal of identifying snow avalanches. Performance was assessed for an optimized model over a range of test data.
- 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
- 11 December 2025
- Published by Wiley
Tracking a rangeland fire with infrasound arrays
We present the first infrasound analyses of a rangeland fire and demonstrate infrasound’s potential in monitoring wildland fires. We recorded a prescribed fire in sagebrush steppe in cool, moist conditions, using eight infrasound arrays of diverse size (3-44 sensors, 10-200 m width) and distance (0-20 km), including arrays inside the burn area. Infrasound arrays within 2 km of fires detected coherent infrasound between 1-10 Hz, both during active ignitions and also while existing fires continued burning during breaks in ignitions. During periods of active ignitions, the direction of arrival estimated by a 44-sensor array closely tracked the actual locations where the helicopter was igniting vegetation. Despite being farther from the fire than four smaller arrays, the 44-sensor array performed best at identifying coherent infrasound and yielding consistent directions of arrival.
- 11 December 2025
- Published by Wiley
Open AccessTracking a rangeland fire with infrasound arrays
We present the first infrasound analyses of a rangeland fire and demonstrate infrasound’s potential in monitoring wildland fires. We recorded a prescribed fire in sagebrush steppe in cool, moist conditions, using eight infrasound arrays of diverse size (3-44 sensors, 10-200 m width) and distance (0-20 km), including arrays inside the burn area. Infrasound arrays within 2 km of fires detected coherent infrasound between 1-10 Hz, both during active ignitions and also while existing fires continued burning during breaks in ignitions. During periods of active ignitions, the direction of arrival estimated by a 44-sensor array closely tracked the actual locations where the helicopter was igniting vegetation. Despite being farther from the fire than four smaller arrays, the 44-sensor array performed best at identifying coherent infrasound and yielding consistent directions of arrival.
- 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.
- 23 April 2025
- Published by Boise State University, Albertsons Library
Dataset for 'A Deep Learning Approach for Infrasound Signal Detection for Snow-covered Areas'
Waveform and statistical data accompanying the journal article "A Deep Learning Approach for Infrasound Signal Detection in Snow-Covered Areas," submitted for consideration to Applied Acoustics. The infrasound sensor array technique is an established method for detecting and monitoring rapid gravity-driven mass movements from infrasound waveform. This technique requires the deployment of two or more sensors at an infrasound station. By employing the array techniques, signals are isolated from various sources of noise by computing cross-correlation measures between recorded signals across spatially distributed sensors at a station. Using multiple sensors requires additional processing that potentially increases event alert time. In this study, we present a signal detection method utilizing deep learning to identify high-quality signals in infrasound waveforms with data from a single infrasound sensor in contrast to multiple sensors required by the array method. This method aims to eradicate the additional processing required by the array techniques. The process involves using audio signal processing techniques to extract discriminatory features from the waveforms, which are then used to train a deep learning model. Data collected over a month period from Little Cottonwood Canyon, Utah, USA was used in this study. The location provides a continuous dataset featuring diverse events that include potential avalanches. Our research is motivated by the goal of identifying snow avalanche signals using a single infrasound sensor. The first step demonstrated in this work involves the use of deep learning techniques to identify high quality signals in recorded infrasound waveforms, including potential avalanche signals. The performance of the resulting optimal model was evaluated across a temporal and spatial scale, consistently achieving a classifier effectiveness score of over 80%.
- 23 April 2025
- Published by Boise State University, Albertsons Library
Dataset for 'A Deep Learning Approach for Infrasound Signal Detection for Snow-covered Areas'
Waveform and statistical data accompanying the journal article "A Deep Learning Approach for Infrasound Signal Detection in Snow-Covered Areas," submitted for consideration to Applied Acoustics. The infrasound sensor array technique is an established method for detecting and monitoring rapid gravity-driven mass movements from infrasound waveform. This technique requires the deployment of two or more sensors at an infrasound station. By employing the array techniques, signals are isolated from various sources of noise by computing cross-correlation measures between recorded signals across spatially distributed sensors at a station. Using multiple sensors requires additional processing that potentially increases event alert time. In this study, we present a signal detection method utilizing deep learning to identify high-quality signals in infrasound waveforms with data from a single infrasound sensor in contrast to multiple sensors required by the array method. This method aims to eradicate the additional processing required by the array techniques. The process involves using audio signal processing techniques to extract discriminatory features from the waveforms, which are then used to train a deep learning model. Data collected over a month period from Little Cottonwood Canyon, Utah, USA was used in this study. The location provides a continuous dataset featuring diverse events that include potential avalanches. Our research is motivated by the goal of identifying snow avalanche signals using a single infrasound sensor. The first step demonstrated in this work involves the use of deep learning techniques to identify high quality signals in recorded infrasound waveforms, including potential avalanche signals. The performance of the resulting optimal model was evaluated across a temporal and spatial scale, consistently achieving a classifier effectiveness score of over 80%.
- 29 July 2024
- Published by Boise State University, Albertsons Library
Dataset for Proximal Analysis of Bonfire Infrasound (Fire Infrasound)
Waveform and video data are provided as an accompaniment to the journal article entitled "Proximal Analysis of Bonfire Infrasound" submitted for consideration at Geophysical Research Letters.
- 29 July 2024
- Published by Boise State University, Albertsons Library
Dataset for Proximal Analysis of Bonfire Infrasound (Fire Infrasound)
Waveform and video data are provided as an accompaniment to the journal article entitled "Proximal Analysis of Bonfire Infrasound" submitted for consideration at Geophysical Research Letters.
- 13 June 2024
- Published by Boise State University, Albertsons Library
Dataset for Induced Resonance and Forces Imparted on Buildings from Volcano Infrasound
Fuego, an active volcano in Guatemala, explodes regularly and generates powerful infrasound whose amplitude regularly exceeds 10 Pa at 10 km where rural communities are located. This study focuses on infrasound recorded at the Fuego Volcano Observatory, a ~250 m3 building located ~8 km from the vent. Records from infrasound sensors situated on either side of a wall facing the volcano were used to measure differential dynamic pressures of more than 50 Pa imparted on the wall facing the volcano. We calculated the building’s infrasonic response as a function of frequency for situations where interior doors were both open and closed and found a remarkably well-defined ~7 Hz tone induced by acoustic oscillations when the interior door was open. The transfer functions, from a week-long period of recording and 34 explosions, provide insight about how buildings could be constructed or sealed to minimize damage such as broken windows.
- 13 June 2024
- Published by Boise State University, Albertsons Library
Dataset for Induced Resonance and Forces Imparted on Buildings from Volcano Infrasound
Fuego, an active volcano in Guatemala, explodes regularly and generates powerful infrasound whose amplitude regularly exceeds 10 Pa at 10 km where rural communities are located. This study focuses on infrasound recorded at the Fuego Volcano Observatory, a ~250 m3 building located ~8 km from the vent. Records from infrasound sensors situated on either side of a wall facing the volcano were used to measure differential dynamic pressures of more than 50 Pa imparted on the wall facing the volcano. We calculated the building’s infrasonic response as a function of frequency for situations where interior doors were both open and closed and found a remarkably well-defined ~7 Hz tone induced by acoustic oscillations when the interior door was open. The transfer functions, from a week-long period of recording and 34 explosions, provide insight about how buildings could be constructed or sealed to minimize damage such as broken windows.