Time filter

Source Type

Bicego M.,University of Verona | Londono-Bonilla J.M.,Observatorio Vulcanologico y Sismologico de Manizales | Orozco-Alzate M.,National University of Colombia
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) | Year: 2015

In this paper we propose a novel framework for the classification of volcano-seismic events, based on strategies and concepts typically employed to classify documents – subsequently largely employed also in other fields. In the proposed approach, we define a dictionary of “seismic words”, used to represent a seismic event as a “seismic document” (i.e. a collection of seismic words). Given this representation, we exploit two well-known models for documents (Bag-of-words and topic models) to derive signatures for seismic events, usable for classification. An empirical evaluation, based on a set of seismic signals from Galeras volcano in Colombia, confirms the potentialities of the proposed scheme, both in terms of interpretability and classification accuracies, also in comparison with standard approaches. © Springer International Publishing Switzerland 2015.

Laeger K.,University of Kiel | Halama R.,University of Kiel | Halama R.,University of Potsdam | Hansteen T.,Leibniz Institute of Marine Science | And 4 more authors.
Journal of South American Earth Sciences | Year: 2013

The last known eruption at Cerro Machín Volcano (CMV) in the Central Cordillera of Colombia occurred ~900yearsBP and ended with the formation of a dacitic lava dome. The dome rocks contain both normally and reversely zoned plagioclase (An24-54), unzoned and reversely zoned amphiboles of dominantly tschermakite and pargasite/magnesio-hastingsite composition and olivine xenocrysts (Fo=85-88) with amphibole/clinopyroxene overgrowth, all suggesting interaction with mafic magma at depth. Plagioclase additionally exhibits complex oscillatory zoning patterns reflecting repeated replenishment, fractionation and changes in intrinsic conditions in the magma reservoir. Unzoned amphiboles and cores of the reversely zoned amphiboles give identical crystallization conditions of 910±30°C and 360±70MPa, corresponding to a depth of about 13±2km, at moderately oxidized conditions (fO2=+0.5±0.2δNNO). The water content in the melt, calculated based on amphibole chemistry, is 7.1±0.4wt.%. Rims of the reversely zoned amphiboles are relatively enriched in MgO and yield higher crystallization temperatures (T=970±25°C), slightly lower melt H2O contents (6.1±0.7wt.%) and overlapping pressures (410±100MPa). We suggest that these rims crystallized following an influx of mafic melt into a resident magma reservoir at mid-crustal depths, further supported by the occurrence of xenocrystic olivine. Crystallization of biotite, albite-rich plagioclase and quartz occurred at comparatively low temperatures (probably <800°C) during early stages of ascent or storage at shallower levels. Based on amphibole mineral chemistry, the felsic resident melt had a rhyolitic composition (71±2wt.% SiO2), whereas the hybrid magma, from which the amphibole rims crystallized, was dacitic (64±3wt.% SiO2). The bulk rock chemistry of the CMV lava dome dacites is homogenous. They have elevated (La/Nb)N ratios of 3.8-4.5, typical for convergent margin magmas, and display several geochemical characteristics of adakites. Both Sr and Nd isotope compositions (87Sr/86Sr ~0.70497, 143Nd/144Nd ~0.51267) are among the most radiogenic observed for the Northern Volcanic Zone of the Andes. They are distinct from oceanic crust that has been subducted in the region, pointing to a continental crustal control on the isotope composition and hence the adakitic signature, possibly in a crustal "hot zone". © 2013 Elsevier Ltd.

Bicego M.,University of Verona | Acosta-Munoz C.,Observatorio Vulcanologico y Sismologico de Manizales | Orozco-Alzate M.,National University of Colombia
IEEE Transactions on Geoscience and Remote Sensing | Year: 2013

The automated classification of seismic volcanic signals has been faced with several different pattern recognition approaches. Among them, hidden Markov models (HMMs) have been advocated as a cost-effective option having the advantages of a straightforward Bayesian interpretation and the capacity of dealing with seismic sequences of different lengths. In the volcano seismology scenario, HMM-based classification schemes were only based on a standard and purely generative scheme, i.e., the Bayes rule: training an HMM per class and classifying an incoming seismic signal according to the class whose model shows the highest likelihood. In this paper, a novel HMM-based classification approach for pretriggered seismic volcanic signals is proposed. The main idea is to enrich the classical HMM scheme with a discriminative step that is able to recover from situations when the classical Bayes classification rule is not sufficient. More in detail, a generative embedding scheme is used, which employs the models to map the signals into a vector space, which is called generative embedding space. In such a space, any discriminative vector-based classifier can be applied. A thorough set of experiments, which is carried out on pretriggered signals recorded at Galeras Volcano in Colombia, shows that the proposed approach typically outperforms standard HMM-based classification schemes, also in some cross-station cases. © 1980-2012 IEEE.

Duin R.P.W.,Technical University of Delft | Orozco-Alzate M.,National University of Colombia | Londono-Bonilla J.M.,Observatorio Vulcanologico y Sismologico de Manizales
Proceedings - International Conference on Pattern Recognition | Year: 2010

Seismic events in and around volcanos, like tremors, earth quakes, ice quakes and strokes of lightning, are usually observed by multiple stations. The question rises whether classifiers trained for one seismic station can be used for classifying observations by other stations, and, moreover, whether a combination of station signals improves the classification performances for a single station. We study this for seismic time signals represented by spectra and spectrograms obtained from 5 seismic stations on the Nevado del Ruiz in Colombia. © 2010 IEEE.

Murcia H.F.,Observatorio Vulcanologico y Sismologico de Manizales | Murcia H.F.,National Autonomous University of Mexico | Sheridan M.F.,State University of New York at Buffalo | Macias J.L.,National Autonomous University of Mexico | Cortes G.P.,Observatorio Vulcanologico y Sismologico de Manizales
Journal of South American Earth Sciences | Year: 2010

Cerro Machín is a dacitic tuff ring located in the central part of the Colombian Andes. It lies at the southern end of the Cerro Bravo-Cerro Machín volcanic belt. This volcano has experienced at least six major explosive eruptions during the last 5000 years. These eruptions have generated pyroclastic flows associated with Plinian activity that have traveled up to 8 km from the crater, and pyroclastic flows associated with Vulcanian activity with shorter runouts of 5 km from the source. Today, some 21,000 people live within a 8 km radius of Cerro Machín. The volcano is active with fumaroles and has shown increasing seismic activity since 2004, and therefore represents a potentially increasing threat to the local population. To evaluate the possible effects of future eruptions that may generate pyroclastic density currents controlled by granular flow dynamics we performed flow simulations with the TITAN2D code. These simulations were run in all directions around the volcano, using the input parameters of the largest eruption reported. The results show that an eruption of 0.3 km3 of pyroclastic flows from a collapsing Plinian column would travel up to 9 km from the vent, emplacing a deposit thicker than 60 m within the Toche River valley. Deposits >45 m thick can be expected in the valleys of San Juan, Santa Marta, and Azufral creeks, while 30 m thick deposits could accumulate within the drainages of the Tochecito, Bermellón, and Coello Rivers. A minimum area of 56 km2 could be affected directly by this kind of eruption. In comparison, Vulcanian column-collapse pyroclastic flows of 0.1 km3 would travel up to 6 km from the vent depositing >45 m thick debris inside the Toche River valley and more than 30 m inside the valleys of San Juan, Santa Marta, and Azufral creeks. The minimum area that could be affected directly by this kind of eruption is 33 km2. The distribution and thickness of the deposits obtained by these simulations are consistent with the hazard map presented by INGEOMINAS (Geological Survey of Colombia) in 2002. The composite map of the simulated flow deposits suggests that after major explosive events such as these, the generation of lahars is probable. © 2009 Elsevier Ltd. All rights reserved.

Discover hidden collaborations