We are IntechOpen,

the world’s leading publisher of

Open Access books
Built by scientists, for scientists

7,500
Open access books available
196,000
International authors and editors
215M
Downloads

Our authors are among the

154
Countries delivered to
TOP 1%
most cited scientists
14%
Contributors from top 500 universities

Selection of our books indexed in the Book Citation Index

in Web of Science™ Core Collection (BKCI)

Interested in publishing with us?

Contact [Link]@[Link]
Numbers displayed above are based on latest data collected.
For more information visit [Link]
Chapter

Machine Learning in Volcanology:

A Review
Roberto Carniel and Silvina Raquel Guzmán

Abstract

A volcano is a complex system, and the characterization of its state at any given
time is not an easy task. Monitoring data can be used to estimate the probability
of an unrest and/or an eruption episode. These can include seismic, magnetic,
electromagnetic, deformation, infrasonic, thermal, geochemical data or, in an ideal
situation, a combination of them. Merging data of different origins is a non-trivial
task, and often even extracting few relevant and information-rich parameters from
a homogeneous time series is already challenging. The key to the characteriza-
tion of volcanic regimes is in fact a process of data reduction that should produce
a relatively small vector of features. The next step is the interpretation of the
resulting features, through the recognition of similar vectors and for example,
their association to a given state of the volcano. This can lead in turn to highlight
possible precursors of unrests and eruptions. This final step can benefit from the
application of machine learning techniques, that are able to process big data in an
efficient way. Other applications of machine learning in volcanology include the
analysis and classification of geological, geochemical and petrological “static” data
to infer for example, the possible source and mechanism of observed deposits, the
analysis of satellite imagery to quickly classify vast regions difficult to investigate
on the ground or, again, to detect changes that could indicate an unrest. Moreover,
the use of machine learning is gaining importance in other areas of volcanology,
not only for monitoring purposes but for differentiating particular geochemical
patterns, stratigraphic issues, differentiating morphological patterns of volcanic
edifices, or to assess spatial distribution of volcanoes. Machine learning is helpful
in the discrimination of magmatic complexes, in distinguishing tectonic settings of
volcanic rocks, in the evaluation of correlations of volcanic units, being particularly
helpful in tephrochronology, etc. In this chapter we will review the relevant meth-
ods and results published in the last decades using machine learning in volcanology,
both with respect to the choice of the optimal feature vectors and to their subse-
quent classification, taking into account both the unsupervised and the supervised
approaches.

Keywords: machine learning, volcano seismology, volcano geophysics,

volcano geochemistry, volcano geology, data reduction, feature vectors

1. Introduction

Pyroclastic density currents, debris flow avalanches, lahars, ash falls can affect
dramatically the life of people living close to volcanoes, and other volcanic products
such as lava flows can severely affect properties and infrastructures. Several volcanoes

1
Updates in Volcanology – Transdisciplinary Nature of Volcano Science

lie close to highly populated areas and the impact of their eruptions could be eco-
nomically very strong. Stochastic forecasts of volcanic eruptions are difficult [1, 2],
but deterministic forecasts (i.e., specifying when, where, how an eruption will occur)
are even harder. Many volcanoes are monitored by observatories that try to estimate
at least the probability of the different hazardous volcanic events [3]. Different time
series can be monitored and hopefully used for forecasting, including seismic data
[4], geomagnetic and electromagnetic data [5], geochemical data [6], deformation
data [7], infrasonic data [8], gas data [9], thermal data from satellite [10] and from
the ground [11]. Whenever possible, a multiparametric approach is always advis-
able. For instance, at Merapi volcano, seismic, satellite radar, ground geodetic and
geochemical data were efficiently integrated to study the major 2010 eruption [12];
a multiparametric approach is essential to understand shallow processes such as the
ones seen at geothermal systems like e.g., Dallol in Ethiopia [13]. Although many time
series may be available, seismic data remain always at the heart of any monitoring
system, and should always include the analysis of continuous volcanic tremor [14];
tremor has in fact a great potential [15] due to its persistence and memory [1, 2] and
its sensitivity to external triggering such as regional tectonic events [16] or Earth tides
[17]. Moreover, its time evolution can be indicative of variations in other parameters,
such as gas flux [18]. Other information-rich time series can be built looking at the
time evolution of the number of the different discrete volcano-seismic events that
can be recorded on a volcano. These include volcano-tectonic (VT) earthquakes,
rockfall events, long-period (LP) and very-long-period (VLP) events, explosions,
etc. Counting the overall number of events is not enough: one has to detect them and
classify them, because they are linked to different processes, as detailed below. For
this reason it is important to generate automatically different time series for each type
of volcano-seismic event.
VT can be described as “normal” earthquakes which take place in a volcanic
environment and can indicate magma movement [19, 20]. LP events have a great
potential for forecasting [21]. Their debated interpretation involves the repeated
expansion and compression of sub-horizontal cracks filled with steam or other
ash-laden gas [22], stick–slip magma motion [23], fluid-driven flow [24], eddy
shedding, turbulent slug flow, soda bottle analogues [25], deformation acceleration
of solidified domes [26] and slow ruptures [27]. Explosion quakes are generated
by sudden magma, ash, and gas extrusion in an explosive event, often associated
to VLP events [28]. In many papers also “Tremor episodes” (TRE events) are
described and counted, usually associated to magma degassing [20]. However, a
volcano with any activity produces a continuous “tremor” which detectability only
depends on the seismic instrumentation sensitivity [29, 30]. So, the class “TRE”
should be better defined as “tremor episode that exceeds the detection limits”. Of
course, at volcanoes we can also record natural but non-volcanic seismic signals
such as far tectonic earthquakes, far explosions, etc., and also anthropogenic signals
e.g., due to industries, ground vehicles, helicopters used for monitoring, etc.
Most volcano observatories rely on manual classification and counting of such
seismic events, which suffers from human subjectivity and can become unfeasible
during an unrest or a seismic crisis [31, 32]. For this reason, manual classification
should be substituted by an automated processing, and here is where machine
learning (ML) comes into place. The same reasoning applies of course also to the
automated processing of other monitoring time series, such as deformation, gas and
water geochemistry, etc. Moreover, ML in volcanology is not restricted to monitoring
active volcanoes but has demonstrated to be useful also when dealing with other
large datasets. Examples include correlating volcanic units in general e.g., [33],
of tephra e.g., [34, 35] and ignimbrites e.g., [36], a task which may become very
difficult especially when many deposits of similar ages and geochemical and

2
Machine Learning in Volcanology: A Review
DOI: [Link]

petrographic characteristics crop out in a given area. ML is also effective for dis-
criminating tectonic settings of volcanic rocks [34, 37]. Recently it has been used
also for the prediction of trace elements in volcanic rocks [38].

2. Machine learning

ML is a field of computer science dedicated to the development of algorithms

which are based on a collection of examples of some phenomenon. These examples
can be natural, human-generated or computer-generated. From another point of
view, ML can be seen as the process of solving a problem by building and using
a statistical model based on an existing dataset [39]. ML can also be defined as
the study of algorithms that allow computer programs to automatically improve
through experience [40]. ML is only one of the ways we expect to achieve Artificial
Intelligence (AI). AI has in fact a wider, dynamic and fuzzier definition, e.g.,
Andrew Moore, former Dean of the School of Computer Science at Carnegie Mellon
University, defined it as “the science and engineering of making computers behave
in ways that, until recently, we thought required human intelligence”. ML is usually
characterized by a series of steps: data reduction, model training, model evaluation,
model final deployment for classification of new, unknown data (see Figure 1). The
training (which is the proper learning phase) can be supervised, semi-supervised,
unsupervised or based on reinforcement.
More data does not necessarily imply better results. Low quality and irrelevant data
can instead lead to worse classification performances. If for each datum we have a very
high number of columns, we may wonder how many of those are really informative.
A number of techniques can help us with this process of data reduction. The simplest
include column variance estimations and evaluating correlations between columns.
Each of the components of the vector that “survive” this phase is called a feature and is
supposed to describe somehow the data item, hopefully in a way that makes it easier to
associate the item to a given class. There are dimensionality reduction algorithms [41]
where the output is a simplified feature vector that is (almost) equally good at describ-
ing the data. There are many techniques to find a smaller number of independent
features, such as Independent Component Analysis (ICA) [42], Non-negative Matrix

Figure 1.
ML can be divided in several steps, from top to bottom. Raw data have first to be reduced by extracting short
and information-rich feature vectors. These can then be used to build models that are trained, analyzed and
finally used for classification of new data. The [labels] are present only in a (semi-)supervised approach.

3
Updates in Volcanology – Transdisciplinary Nature of Volcano Science

Factorization (NMF) [43], Singular Value Decomposition [44], Principal Component

Analysis (PCA) [45] and Auto-encoders [46]. Linear Discriminant Analysis (LDA)
[47] uses the training samples to estimate the between-class and within-class scatter
matrices, and then employs the Fisher criterion to obtain the projection matrix for
feature extraction (or feature reduction).
In supervised learning, the dataset is a collection of example couples of the
type (data, label) {( xi ,yi )}i =1.. N . Each element xi is called a feature vector and has a
companion label yi . In the supervised learning approach the dataset is used to
derive a model that takes a feature vector as input and outputs a label that should
describe it. For example, the feature vector of volcano-seismic data could contain
several amplitude-based, spectral-based, shape-based or dynamical parameters and
the label to be assigned could be one of those described above, i.e., VT, LP, VLP. In a
volcanic geochemical example, feature vectors could contain major elements weight
percentages, and labels the corresponding rock type. The reliability of the labels is
often the most critical issue of the setup of a supervised ML classification scheme.
Labels should therefore be assigned carefully by experts. In general, it is much
better to have relatively few training events with reliable labels than to have many
more, but not so reliable, labeled examples.
In unsupervised learning, the dataset is a collection of examples without any
labeling, i.e., containing only the data { xi }i =1.. N . As in the previous case, each xi is
a feature vector, and the goal is to create a model that maps a feature vector x into a
value (or another vector) that can help solving a problem. Typical examples are all
the clustering procedures, where the output is the cluster number to which each
datum belongs. The choice of the best features to use is a difficult one, and several
techniques of Unsupervised Feature Selection were proposed, with the capability of
identifying and selecting relevant features in unlabeled data [48]. Unsupervised
outlier detection methods [49] can also be used, where the output indicates if a
given feature vector is likely to describe a “normal” or “anomalous” member of the
dataset.
The semi-supervised learning approach stands somehow in the middle, and
the dataset contains both labeled (usually a few) and unlabeled (usually many
more) feature vectors. The basic idea is similar to supervised learning, but with the
possibility to exploit also the presence of (many more) unlabeled examples in the
training phase.
In reinforcement learning, the machine is “embedded” in an environment,
which state is again described by a feature vector. In each state the machine can
execute actions, which produce different rewards and can cause an environmental
state transition. The goal in this case is to learn a policy, i.e., a function or model
that takes the feature vector as input and outputs an optimal action to execute
in that state. The action is optimal if it maximizes the expected average reward.
We can also say that reinforcement learning is a behavioral learning model. The
algorithm receives feedback from the data analysis, guiding the user to the best
outcome. Here the main point is that the system is not trained with a sample
dataset but learns through trial and error. Therefore, a sequence of successful
decisions will result in that process being reinforced, because it best solves the
problem at hand. Problems that can be tackled with this approach are the ones
where decision making is sequential, and the goal is long-term, such as game
playing, robotics, resource management, or logistics. Time is therefore explicitly
used here, contrary to other approaches, in which in most of the cases data items
are analyzed one by one without taking into account the time order in which
they arrive.
In some domains (and volcanology is a good example) training data are scarce.
In this case we can profit from knowledge acquired in another domain using

4
Machine Learning in Volcanology: A Review
DOI: [Link]

techniques known as Transfer Learning (TL) [50]. The basic idea here is to train
a model in one domain with abundant data (original domain) and then use it as
a pretrained model in a different domain (with less data). There is a successive
fine-tuning phase using domain-specific available data (in the target domain). This
approach was applied for instance at Volcán de Fuego de Colima (Mexico) [51],
Mount St. Helens (USA) and Bezymianny (Russia) [52].
Among the computer languages that are most used for implementing ML
techniques we can cite Python [53], R [54], Java [55], Javascript [56], Julia [57] and
Scala [58]. Many dedicated, open source libraries are available for each of them, and
many computer codes, also specialized for volcanic and geophysical data, can be
found in open access repositories such as GitHub [59].

3. Machine learning techniques

Extracted feature vectors can become inputs to several different techniques

of machine learning. We can cite among others Cluster Analysis (CA) [60],
Self-Organizing Maps (SOM) [61–63], Artificial Neural Networks (ANN) and
Multi Layer Perceptrons (MLP) [64–66], Support Vector Machines (SVM) [67],
Convolutional Neural Networks (CNN) [51], Recurrent Neural Networks (RNN)
[68], Hidden Markov Models (HMM) [3, 31, 69–71] and their Parallel System
Architecture (PSA) based on Gaussian Mixture Models (GMM) [72].
CA (Figure 2a) is an unsupervised learning approach aimed at grouping similar
data while separating different ones, where similarity is measured quantitatively
using a distance function in the space of feature vectors. The clustering algorithms
can be divided into hierarchical and non-hierarchical. In the former a tree-like
structure is built to represent the relations between clusters, while in the latter new
clusters are formed by merging or splitting existing ones without following a tree-
like structure but just grouping the data in order to maximize or minimize some
evaluation criteria. CA includes a vast class of algorithms, including e.g., K-means,
K-medians, Mean-shift, DBSCAN, Expectation–Maximization (EM), Clustering
using Gaussian Mixture Models (GMM), Agglomerative Hierarchical, Affinity
Propagation, Spectral Clustering, Ward, Birch, etc. Most of these methods are
described and implemented in the open-source Python package scikit-learn [73].
The use of six different unsupervised, clustering-based methods to classify volcano
seismic events was explored at Cotopaxi Volcano [32]. One of the most difficult
issues is the choice of the number of clusters into which the data should be divided;
this number in most of the cases has in fact to be fixed a priori before running
the code. Several techniques exist in order to help with this choice, such as elbow,
silhouette, gap statistics, heuristics, etc. Many of them are described and included
in the R package NbClust [74]. Problems arise when the estimates that each of them
provides are contradictory.
Another approach to unsupervised classification is SOM (Figure 2b) or
Kohonen maps [75, 76], a type of ANN trained to produce a low dimensional,
usually 2D, discretized representation of the feature vector space. The training is
based on competitive and collaboration learning, using a neighborhood function to
preserve the input topological properties.
A very common type of ANN, often used for supervised classification, is MLP,
which consists of at least three layers of nodes (Figure 2c): an input layer, (at least)
one hidden layer and an output layer. Nodes use nonlinear activation functions and
are trained through the backpropagation mechanism. If the number of hidden layers
of an ANN becomes very high, we talk of Deep Neural Networks (DNN), which are
also used mainly in a supervised fashion. Among DNN, the CNN (Figure 2d) contain

5
Updates in Volcanology – Transdisciplinary Nature of Volcano Science

Figure 2.
Schematic illustration of some of the ML techniques described in the text. (a) Cluster analysis in its hierarchical
and non-hierarchical versions. (b) Self-organizing maps (c) multilayer perceptron (d) convolutional neural
network.

at least some convolutional layers, that convolve their inputs with a multiplication or
other dot product. The activation function in the case of CNN is commonly a recti-
fied linear unit (ReLU), and there are also pooling layers, fully connected layers and
normalization layers.
A RNN is a type of ANN with a feedback loop (Figure 3a), in which neuron
outputs can also be used as neuron inputs in the same layer, allowing to maintain
some information during the training process. Long Short Term Memory networks
(LSTM) are a subset of RNN, capable of learning long-term dependencies [77] and
better remember information for long periods of time. RNN can be used for both
supervised and unsupervised learning.

6
Machine Learning in Volcanology: A Review
DOI: [Link]

Figure 3.
Schematic illustration of some of the ML techniques described in the text. (a) Recurrent neural network
(b) logistic regression (c) support vector machine (d) random forest (e) hidden Markov model.

Logistic regression (LR) (Figure 3b) is a supervised generalized linear model,

i.e., the classification (probability) dependence on the features is linear [78]. In order
to avoid the problems linked to high dimensionality of the data, techniques such as
the Least Absolute Shrinkage and Selection Operator (LASSO) can be applied to
reduce the number of dimensions of the feature vectors which are input to LR [79].
SVM (Figure 3c) constitute a supervised statistical learning framework [80]. It
is most commonly used as a non-probabilistic binary classifier. Examples are seen as
points in space, and the aim is to separate categories by a gap that is as wide as possible.
Unknown samples are then assigned to a category based on the side of the gap on
which they fall. In order to perform a non-linear classification, data are mapped into
high-dimensional feature spaces using suitable kernel functions.

7
Updates in Volcanology – Transdisciplinary Nature of Volcano Science

Sparse Multinomial Logistic Regression (SMLR) is a class of supervised methods

for learning sparse classifiers that incorporate weighted sums of basis functions
with sparsity-promoting priors encouraging the weight estimates to be either
significantly large or exactly zero [81]. The sparsity concept is similar to the one at
the base of Non-negative Matrix Factorization (NMF) [82]. The sparsity-promoting
priors result in an automatic feature selection, enabling to somehow avoid the
so-called “curse of dimensionality”. So, sparsity in the kernel basis functions and
automatic feature selection can be achieved at the same time [83]. SMLR methods
control the capacity of the learned classifier by minimizing the number of basis
functions used, resulting in better generalization. There are fast algorithms for
SMLR that scale favorably in both the number of training samples and the feature
dimensionality, making them applicable even to large data sets in high-dimensional
feature spaces.
A Decision Tree (DT) is an acyclic graph. At each branching node, a specific
feature xi is examined. The left or right branch is followed depending on the value
of xi in relation to a given threshold. A class is assigned to each datum when a leaf
node is reached. As usual, a DT can be learned from labeled data, using different
strategies. In the DT class we can mention Best First Decision Tree (BFT),
Functional Tree (FT), J48 Decision Tree (J48DT), Naïve Bayes Tree (NBT) and
Reduced Error Pruning Trees (REPT). Ensemble learning techniques such as
Random SubSpace (RSS) can be used to combine the results of the different
trees [84].
The Boosting concept, a kind of ensemble meta-algorithm mostly (but not only)
associated to supervised learning, uses original training data to create iteratively
multiple models by using a weak learner. Each model would be different from the
previous one as the weak learners try to “fix” the errors made by previous models.
An ensemble model will then combine the results of the different weak models. On
the other side, Bootstrap aggregating, also called by the contracted name Bagging,
consists of creating many “almost-copies” of the training data (each copy is slightly
different from the others) and then apply a weak learner to each copy and finally
combine the results. A popular and effective algorithm based on bagging is Random
Forest (RF). Random Forest (Figure 3d) is different from the standard bagging in
just one way. At each learning step, a random subset of the features is chosen; this
helps to minimize correlation of the trees, as correlated predictors are not efficient
in improving classification accuracy. Particular attention has to be taken in order to
best choose the number of trees and the size of the random feature subsets.
A Hidden Markov Model (HMM) (Figure 2e) is a statistical model in which the
system being modeled is assumed to be a Markov process. It describes a sequence
of possible events for which the probability of each event depends only on the state
occupied in the previous event. The states are unobservable (“hidden”) but at each
state the Model emits a “message” which depends probabilistically on the current
state. Applications are wide in scope, from reinforcement learning to temporal pat-
tern recognition, and the approach works well when time is important; speech [85],
handwriting and gesture recognition are then typical fields of applications, but also
volcano seismology [69, 86].

4. Applications to seismo-volcanic data

Eruptions are usually preceded by some kind of change in seismicity, making

seismic data one of the key dataset in any attempt to forecast volcanic activity [4].
As we mentioned before, manual detection and classification of discrete events
can be very time consuming, up to becoming unfeasible during a volcanic crisis.

8
Machine Learning in Volcanology: A Review
DOI: [Link]

An automatic classification procedure becomes therefore highly valuable, also as a

first step towards forecasting techniques such as material Failure Forecast Method
(FFM) [87, 88]. Feature vectors should be built in order to provide most informa-
tion about the source, minimizing e.g., path and site effects. In many cases features
can be independent from a specific physical model describing a phenomenon. This
allows ML to work well even when there is no scientific agreement on the generation
of a given seismic signal. A good example in volcano seismology is given by the LP
events. Standardizing data, making them independent from unwanted variables
is also in general a convenient approach [31]. Time-domain and spectral-based
amplitudes, spectral phases, auto- and cross-correlations, statistical and dynamical
parameters have been considered as the output of data reduction procedures that
can be included into feature vectors [14]. In the literature, these have included linear
predictor coding for spectrograms [66], wavelet transforms [89], spectral autocor-
relation functions [90], statistical and cepstral coefficients [91]. Extracted feature
vectors become then the input to one or another ML method.
CA is probably the most used class of unsupervised techniques and the applica-
tions to volcano seismology follow this general rule. Spectral clustering was applied
e.g., to seismic data of Piton de la Fournaise [60]. The fact that e.g., LP seismic
signals can be clustered into families indicates that the family members are very
similar to each other. The existence of similar events implies similar location and
similar source process, i.e., it means the presence of a source that repeats over time in
an almost identical way. Clustering data after some kind of normalization forces CA
algorithms to look for similar shapes, independently of size. If significant variations
in amplitude are then seen within families, this can indicate that the source processes
of these events are not only repeatable but also scalable in size, as observed e.g., at
Soufrière Hills Volcano, Montserrat [92] or at Irazú, Costa Rica [93]. The similarity
of events in the different classes can then be used to detect other events, e.g., for the
purpose of stacking them and obtain more accurate phase arrivals; this was done
e.g., at Kanlaon, Philippines [94]. For this purpose, an efficient open-source package
is available, called Repeating Earthquake Detector in Python (REDPy) [95].
In volcano-seismology SOM were applied e.g., to Raoul Island, New Zealand
[61]. A hierarchical clustering was applied to results of SOM tremor analysis at
Ruapehu [62] and Tongariro [96] in New Zealand, using the Scilab environment.
A similar combined approach was applied in Matlab to Etna volcanic tremor
[97]. Several geometries of SOM were used, with rectangular or hexagonal
nearest neighbors cells, planar, toroidal or spherical maps, etc. [61]. The clas-
sic ANN/MLP approach was applied e.g., to seismic data recorded at Vesuvius
[66], Stromboli [98], Etna [99], while DNN architectures were applied e.g., to
Volcán de Fuego, Colima [100]. The use of genetic algorithms for the optimiza-
tion of the MLP configuration was proposed for the analysis of seismic data of
Villarrica, Chile [101]. CNN were applied e.g., to Llaima Volcano (Chile) seismic
data, comparing the results to other methods of classification [102]. RNNs were
applied, together with other methods, to classify signals of Deception Island
Volcano, Antarctica [68]. The architectures were trained with data recorded in
1995–2002 and models were tested on data recorded in 2016–2017, showing good
generalization accuracy.
Supervised LR models have been applied in the estimation of landslide suscepti-
bility [103] and to volcano seismic data to estimate the ending date of an eruption at
Telica (Nicaragua) and Nevado del Ruiz (Colombia) [104]. SVM were applied many
times to volcano seismology e.g., to classify volcanic signals recorded at Llaima,
Chile [105] and Ubinas, Peru [106]. Multinomial Logistic Regression was used,
together with other methods, to evaluate the feasibility of earthquake prediction
using 30 years of historical data in Indonesia, also at volcanoes [107].

9
Updates in Volcanology – Transdisciplinary Nature of Volcano Science

RF was applied to the discrimination of rockfalls and VT recorded at Piton de la

Fournaise in 2009–2011 and 2014–2015. 60 features were used, and excellent results
were obtained. However, a RF trained with 2009–2011 data did not perform well on
data recorded in 2014–2015, demonstrating how difficult it is to generalize models
even at the same volcano [108]. RF, together with other methods, was recently used
on volcano seismic data with the specific purpose to determine when an eruption
has ended [104], a problem which is far from being trivial. RF was also used to
derive ensemble mean decision tree predictions of sudden steam-driven eruptions
at Whakaari (New Zealand) [109].
Most of the methods described so far try to classify discrete seismic events that
were already extracted from the continuous stream, i.e., already characterized by a
given start and end. There are therefore in general two separated phases: detection
and classification [106]. Continuous HMM on the other side are able to process
continuous data and can therefore extract and classify in a single, potential real-
time, step. HMM are finite-state machines and model sequential patterns where
time direction is an essential information. This is typical of (volcano) seismic data.
For instance, P waves always arrive before S waves. HMM-based volcanic seismic
data classifiers have therefore been used by many authors [87, 110–113]. HMM are
also used routinely in some volcano observatories e.g., at Colima and Popocatepetl
in Mexico [71]. Etna seismic data was processed by HMM applied to characters
generated by the Symbolic Aggregate approXimation (SAX) which maps seismic
data into symbols of a given alphabet [114]. HMM can be also combined with
standardization procedures such as Empirical Mode Decomposition (EMD) when
classifying volcano seismic data [31].
Another characteristic common to many of the applications published in the
literature is the fact that feature vectors are extracted from data recorded at a
single station. There are relatively few attempts to build multi-station classification
schemes. At Piton de la Fournaise a system based on RF was implemented [115]. At
the same volcano, a multi-station approach was used to classify tremor measure-
ments and identify fundamental frequencies of the tremor associated to different
eruptive behavior [60]. A scalable multi-station, multi-channel classifier, using
also the empirical mode decomposition (EMD) first proposed by [31] was applied
to Ubinas volcano (Peru). The principal component analysis is used to reduce the
dimensionality of the feature vector and a supervised classification is carried out
using various methods, with SVM obtaining the best performance [116]. Of course,
with a multi-station approach particular care has to be taken in order to build a
system which is robust with respect to the loss of one or more seismic stations due to
volcanic activity or technical failures.
Open source software and open access papers are luckily becoming more and
more common. If we consider the processing and classification of volcano seismic
data, several tools are now available for free download and use, especially within
the Python environment. Among the most popular, we can cite ObsPy [117] and
Msnoise [118], with which researchers and observatories can easily process big
quantities of continuous seismic data. Once these tools have produced suitable
feature vectors, we can look for open source software to implement the different ML
approaches described in this contribution. Many generic ML libraries are available
e.g., on GitHub [59] but very few are dedicated specifically to the classification of
volcano seismic data. Among these, we can cite the recent package Python Interface
for the Classification of Seismic Signals (PICOSS) [119]. It is a graphical, modular
open source software for detection, segmentation and classification of seismic data.
Modules are independent and adaptable. The classification is currently based on
two modules that use Frequency Index analysis [120] or a multi-volcano pre-trained
neural network, in a transfer learning fashion [52]. The concept of a multi-volcano

10
Machine Learning in Volcanology: A Review
DOI: [Link]

recognizer is also at the core of the EU-funded [Link] project [31, 121]. The
aim is to build an automatic Volcano Seismic Recognition (VSR) system, conceptu-
ally supervised (as it is based on HMM) but practically unsupervised, because once
it is trained on a number of volcanoes with labeled sample data, it can be used on
volcanoes without any previous data in an unsupervised fashion. The idea is in fact
to build robust models trained on many datasets recorded by different teams on
different volcanoes, and to integrate these models on the routinely used monitor-
ing system of any volcano observatory. Also in this case, the open source software
is made freely available; this includes a command interface called PyVERSO [122]
based on HTK, a speech recognition HMM toolkit [123], a graphical interface called
geoStudio and a script called liveVSR, able to process real-time data downloaded
from any online seismic data server [124], together with some pre-trained ML
models [125].
As we mentioned before, in order to train supervised models for classifying
seismic events, few events with reliable labels are better than many unreliably
labeled examples. Just to give a rough idea, 20 labeled events per class is a good
starting point, but a minimum of 50 labeled events per class is recommended.
Labelling discrete events is enough for many methods, but for approaches like
HMM, where the concept is to run the classification on continuous data, it is
essential to have a sufficient number of continuously labeled time periods, in order
to “show” the classifier enough examples of transition from tremor to a discrete
event, and then back to tremor. It is important to have many examples also of
“garbage” events, i.e., events we are not interested in, so that the classifier can rec-
ognize and discard them. Finally, it is advisable to have a wide variability of events
within each given class rather than having many very similar events. There is not
yet an agreement on a single file format to store these labels. As speech recognition
is much older and more developed than seismic recognition, it is suggested to adopt
standard labelling formats of that domain, i.e., the transcription MLF files, which
are normal text files that include for each event the start time, the end time and of
course the label. These files can be created manually with a simple text editor, or by
using a program with a GUI, such as geoStudio [124] or Seismo_volcanalysis [126].
Other graphical software packages like SWARM [127] use other formats to store
the labels, such as CSV, but it is always possible to build scripts that convert the
resulting label files into MLF format, which remains the recommended one.

5. Applications of machine learning to geochemical data

ML applications to geochemical data of volcanoes are increasing in the last

years, although most of them are limited to the use of cluster analysis. CA has been
used for example to identify and quantify mixing processes using the chemistry of
minerals [128], also for the study of volcanic aquifers [129, 130] or to differentiate
magmatic systems e.g., [131]. Platforms used to carry out these analyses include
the Statistical Toolbox in Matlab [132], or the R platform [54]; some geochemical
software made in this last platform include the CA as the GCDkit [33]. In most ML
analyses on geochemical samples it is common to use whole rock major elements
and selected trace elements; some applications also include isotopic ratios. Many
ML applications to geochemical data use more than one technique, frequently
combining both unsupervised and supervised approaches.
A combination of SVM, RF and SMLR approaches were used by [37] to account
for variations of geochemical composition of rocks from eight different tectonic
settings. The authors note that SVM used to discriminate tectonic settings as used
by [34] is a powerful tool. The RF approach is shown to have the advantage, with

11
Updates in Volcanology – Transdisciplinary Nature of Volcano Science

respect to SVM, of providing the importance of each feature during discrimination.

The weakness of applying the RF for tectonic setting discrimination is that the
evaluation based only on a majority vote of multiple decision trees often makes the
obtained quantitative geochemical interpretation of these elements and isotopic
ratios difficult. The authors suggest that the best quantitative discriminant is that
of SMLR, as it allows to assign to each sample a probability of belonging to a given
group (tectonic setting in this case), with still the possibility of identifying the
importance of each feature. This tool is a notable step forward in the discrimination
of the geochemical signature of the different tectonic settings, which is commonly
assessed based on binary or ternary diagrams e.g., [133, 134] which are useful with
many samples but are not able to differentiate a tectonic setting where a complex
evolution of magmas has occurred. In the last decade multielement variation
diagrams were proposed e.g., [135] and also the use of Decision Trees e.g., [136] or
LDA e.g., [137] to accurately assign a tectonic setting based on rock geochemistry.
Based on rock sample geochemistry, [37] show that a set of 17 elements and isoto-
pic ratios is needed to clearly identify the tectonic setting. Two new discriminant
functions were recently proposed to discriminate the tectonic settings of mid-ocean
ridge (MOR) and oceanic plateau (OP). 10 datasets (original concentrations as well
as isometric log-ratio transformed variables; all 10 major elements as well as all 10
major and 6 trace elements) were used to evaluate the quality of discrimination
from LDA and canonical analysis [138].
The software package Compositional Data Package (CoDaPack) [139] and a
combination of unsupervised (CA) and supervised (LDA) learning approaches
was used by [36] to identify compositional variation of ignimbrite magmas in the
Central Andes, trying to use these methods as a tool for ignimbrite correlation.
They have used the Statistica software [140] for both CA and LDA.
Correlating tephra and identifying their volcanic sources is a very difficult task,
especially in areas where several volcanoes had explosive eruptions in a relatively
short period of time. This is particularly challenging when volcanoes have similar
geochemical and petrographic compositions. Electron microprobe analysis of glass
compositions and whole-rock geochemical analyses are used frequently to make
these correlations. However, correlations may not be so accurate when using only
geochemical tools that may mask diagnostic variability; sometimes one of the most
important advantages of ML in this regard is the speed at which correlations can
be made, rather than the accuracy [35]. Other contributions however demonstrate
how ML techniques can make these correlations also accurate. Some highly accurate
results of ML techniques applied to tephra correlation include those of LDA [141, 142]
and SVM e.g., [143]; however, SVM may fail in specific cases and for the case study
of tephra from Alaska volcanoes, the combination of ANN and RF are the best ML
techniques to apply [35]. The authors use the R software [54] to apply these methods,
and they underline the advantage of producing probabilistic outputs.
SOM was used as an unsupervised neural network approach to analyze geo-
chemical data of Ischia, Vesuvius and Campi Flegrei [144]. The advantage of this
method is that there is no need of previous knowledge of geochemical or petrologi-
cal characteristics and that it allows the use of large databases with large number
of variables. The SOM toolbox for Matlab [132] was used by [144] to perform two
tests, the first based on major elements and selected trace elements to find similar
evolution processes, the second to investigate the magmatic source, so a vector
containing a selection of ratios between major and trace elements was adopted.
One of the enhancements of this method is that the resulting clusters permitted to
differentiate rock samples that were only comparably distinguished by 2D diagrams
of isotopic ratios; in other words, similar results were obtained with the limited
availability of less expensive geochemical data.

12
Machine Learning in Volcanology: A Review
DOI: [Link]

One of the applications of ML techniques that maybe extremely useful in geo-

chemistry is the apparent possibility of predicting the concentration of unknown
elements if a large number of data of other elements is known. A combination of
ML techniques was used by [38] to predict Rare Earth Elements (REE) concentra-
tions on Ocean Island Basalts (OIB) using RF. They used 1283 analyses of which
80% were used for training and the remaining 20% to validate the results. They
found good estimations only in the Light Rare Earth Elements (LREE), suggesting
that the results may be improved by using a larger set of input data for training.
One possible solution may be the use of not only major elements for training but
also of other trace elements obtained through the same analytical method of major
elements.
The origin of the volcanoes in Northeast China, analyzed by RF and DNN using
the full chemical compositional data, was associated to the Pacific slab, subducting
at Japan, reaching ~600-km depth under eastern China, and extending horizontally
up to Mongolia. The boundary between volcanoes triggered by fluids and melts
from the slab and those not related to it was located at the westernmost edge of the
deeply buried Pacific slab [145].
As highlighted by [143] ML methods require the integration with other tech-
niques such as fieldwork, petrographic observations and classic geochemical studies
to obtain a clearer picture of the investigated problem. While in other fields, it is
relatively easy (and cheap) to acquire big amounts of data (hundreds or more), this
is not the case for geochemistry. However, we underline that the application of ML
techniques to the geochemistry of volcanic rocks does need a minimum dataset size.
In the literature a set of 250 analyses is described as sufficiently large amount of
data but, as usual, one can try using the available data (often even less than 50) but
thousands of examples would definitely improve the results.

6. Applications of machine learning to other volcanological data

ML appears more and more often in volcanology literature, and specific fields of
application span now also other sub-disciplines.
Mount Erebus in Antarctica has a persistent lava lake showing Strombolian
activity, but its location is definitely remote. Therefore, automatic methods to
detect these explosions are highly needed. A CNN was trained using infrared
images captured from the crater rim and “labeled” with the help of accompany-
ing seismic data, which was not used anymore during the subsequent automatic
detection [146].
Clast morphology is a fundamental tool also for studies concerning volcanic
textures. Texture analysis of clasts provides in particular information about genesis,
transport and depositional processes. Here, ML has still to be developed fully but
e.g., the application of preprocessing techniques such as the Radon transform can be
a first step towards an efficient definition of feature vectors to be used for classifica-
tion, as shown e.g., at Colima volcano [147].
The Museum of Mineralogy, Petrography and Volcanology of the University of
Catania implemented a communication system based on the visitor’s personal expe-
rience to learn by playing. There is a web application called I-PETER: Interactive
Platform to Experience Tours and Education on the Rocks. This platform includes
a labeled dataset of images of rocks and minerals to be used also for petrological
investigations based on ML [148].
Satellite remote sensing technology is increasingly used for monitoring the sur-
face of the Earth in general, and volcanoes in particular, especially in areas where
ground monitoring is scarce or completely missing. For instance, in Latin America

13
Updates in Volcanology – Transdisciplinary Nature of Volcano Science

202 out of 319 Holocene volcanoes did not have seismic, deformation or gas moni-
toring in 2013 [7]. A complex-valued CNN was proposed to extract areas with land
shapes similar to given samples in interferometric synthetic aperture radar (InSAR),
a technique widely applied in volcano monitoring. An application was presented
grouping similar small volcanoes in Japan [149]. InSAR measurements have great
potential for volcano monitoring, especially where images are freely available. ML
methods can be used for the initial processing of single satellite data. Processing of
potential unrest areas can then fully exploit integrated multi-disciplinary, multi-
satellite datasets [7]. The Copernicus Programme of the European Space Agency
(ESA) and the European Union (EU) has recently contributed by producing the
Sentinel-2 multispectral satellites, able to provide high resolution satellite data
for disaster monitoring, as well as complementing previous satellite images like
Landsat. The free access policy also promotes an increasing use of Sentinel-2 data,
which is often processed by ML techniques such as SVM and RF [150]. A transfer
learning strategy was applied to ground deformation in Sentinel-1 data [151] and a
range of pretrained networks was tested, finding that AlexNet [152] is best suited to
this task. The positive results were checked by a researcher and fed back for model
updating.
The global volcano monitoring platform MOUNTS (Monitoring Unrest from
Space) uses multisensor satellite-based imagery (Sentinel-1 Synthetic Aperture
Radar SAR, Sentinel-2 Short-Wave InfraRed SWIR, Sentinel-5P TROPOMI),
ground-based seismic data (GEOFON and USGS global earthquake catalogs), and
CNN to provide support for volcanic risk assessment. Results are visualized on an
open-access website. The efficiency of the system was tested on several eruptions
(Erta Ale 2017, Fuego 2018, Kilauea 2018, Anak Krakatau 2018, Ambrym 2018, and
Piton de la Fournaise 2018–2019) [153].
Debris flow events are one of the most widespread and dangerous natural
processes not only on volcanoes but more in general in mountainous environ-
ments. A methodology was recently proposed [154] that combines the results of
deterministic and heuristic/probabilistic models for susceptibility assessment. RF
models are extensively used to represent the heuristic/probabilistic component
of the modeling. The case study presented is given by the Changbai Shan volcano,
China [154].
Mapping lava flows from satellite is another important remote sensing applica-
tion. RF was applied to 20 individual flows and 8 groups of flows of similar age
using a Landsat 8 image and a DEM of Nyamuragira (Congo) with 30 m resolution.
Despite spectral similarity, lava flows of contrasting age can be well discriminated
and mapped by means of image classification [155].
The hazard related to landslides at volcanoes is also significant. DNN models
were proposed for landslide susceptibility assessment in Viet Nam, showing consid-
erable better performance with respect to other ML methods such as MLP, SVM, DT
and RF [156]. The use of DNN approach could be therefore an interesting approach
for the landslide susceptibility mapping of active volcanoes.
Muon imaging has been successfully used by geophysicists to investigate the
internal structure of volcanoes, for example at Etna (Italy) [157]. Muon imaging is
essentially an inverse problem and it can profit from the application of ML tech-
niques, such as ANN and CA [158].
Combinations of supervised and unsupervised ML techniques have been used to
map volcanoes also on other planets. A ML paradigm was designed for the iden-
tification of volcanoes on Venus [159]. Other studies have used topographic data,
such as DEM and associated derivatives obtained from orbital images, to detect and
classify manually labeled Martian landforms including volcanoes [160].

14
Machine Learning in Volcanology: A Review
DOI: [Link]

7. Conclusions

ML techniques will have an increasing impact on how we study and model

volcanoes in all their aspects, how we monitor them and how we evaluate their
hazards, both in the short and in the long term. The increasing number of moni-
toring equipment installed on volcanoes on one side provides more and more data,
on the other often causes their real time processing unfeasible especially when
most needed i.e., during unrest and eruptions. Here ML will show its best useful-
ness, as it can provide the perfect tools to sift through big data to identify subtle
patterns that could indicate unrest, hopefully well before eruptions. One impor-
tant issue is the one of generalization. We must go towards the construction of ML
models that can be applied on different volcanoes, for instance when previous data
is not available for training specific models. The concepts of transfer learning can
be important here.
The routine use of ML tools at the different volcano observatories should be
promoted by providing easy installation procedures and easy integration into
existing monitoring systems. Open source software should be always chosen
whenever possible. On the other hand, observatories should provide good open
training data to ML developers, researchers and data scientists in order to improve
the models in a virtuous circle. An easy availability of open access data, both from
the ground and from satellites should be exploited for building reliable training
sets in the different fields of volcanology. This will allow “scientific competition”
between research groups using different ML approaches and make a direct com-
parison of results easier, like it is common in other disciplines where “standard”
training datasets are available for download to everybody.

Acknowledgements

RC wishes to acknowledge the invaluable help resulted from discussions with his
coauthors during previous works; in particular, collaborations with Luca Barbui,
Moritz Beyreuther, Corentin Caudron, Guillermo Cortés, Art Jolly, Philippe Lesage,
Joachim Wassermann.
This review is partially based on the results of a previous project funded under
the European Union’s Horizon 2020 research and innovation program under the
Marie Skłodowska-Curie Grant Agreement No. 749249 ([Link]).

Conflict of interest

The authors declare no conflict of interest.

15
Updates in Volcanology – Transdisciplinary Nature of Volcano Science

Author details

Roberto Carniel1* and Silvina Raquel Guzmán2

1 DPIA, University of Udine, Udine, Friuli, Italy

2 IBIGEO, UNSa-CONICET, Rosario de Lerma, Salta, Argentina

*Address all correspondence to: [Link]@[Link]

© 2020 The Author(s). Licensee IntechOpen. This chapter is distributed under the terms
of the Creative Commons Attribution License ([Link]
by/3.0), which permits unrestricted use, distribution, and reproduction in any medium,
provided the original work is properly cited.

16
Machine Learning in Volcanology: A Review
DOI: [Link]

References

[1] O. Jaquet, R. Carniel, S. Sparks, [8] E. Marchetti et al., “Long range

G. Thompson, R. Namar, and M. Di infrasound monitoring of Etna volcano,”
Cecca, “DEVIN: A forecasting approach Sci. Rep., vol. 9, no. 1, 2019.
using stochastic methods applied to
the Soufrière Hills Volcano,” J. Volcanol. [9] A. Aiuppa et al., “Unusually large
Geotherm. Res., vol. 153, no. 1-2 SPEC. magmatic CO2 gas emissions prior to a
ISS., pp. 97-111, May 2006. basaltic paroxysm,” Geophys. Res. Lett.,
2010.
[2] O. Jaquet and R. Carniel,
“Multivariate stochastic modelling: [10] F. Marchese, N. Pergola, and L.
Towards forecasts of paroxysmal phases Telesca, “Investigating the temporal
at Stromboli,” J. Volcanol. Geotherm. fluctuations in satellite advanced very
Res., vol. 128, no. 1-3, pp. 261-271, Nov. high resolution radiometer thermal
2003. signals measured in the volcanic area of
Etna (Italy),” Fluct. Noise Lett., vol. 6,
[3] W. P. Aspinall, R. Carniel, O. Jaquet, no. 3, pp. L305-L316, 2006.
G. Woo, and T. Hincks, “Using hidden
multi-state Markov models with multi- [11] A. J. L. L. Harris, R. Carniel, and
parameter volcanic data to provide J. Jones, “Identification of variable
empirical evidence for alert level convective regimes at Erta Ale Lava
decision-support,” J. Volcanol. Geotherm. Lake,” J. Volcanol. Geotherm. Res., vol.
Res., vol. 153, no. 1-2 SPEC. ISS., 142, no. 3-4, pp. 207-223, Apr. 2005.
pp. 112-124, 2006.
[12] Surono et al., “The 2010 explosive
eruption of Java’s Merapi volcano-A
[4] R. Ortiz, A. García, J. M. Marrero,
‘100-year’ event,” J. Volcanol. Geotherm.
S. la Cruz-Reyna, R. Carniel, and J.
Res., vol. 241-242, pp. 121-135, 2012.
Vila, “Volcanic and volcano-tectonic
activity forecasting: A review on seismic
[13] R. Carniel, E. M. Jolis, and J. Jones,
approaches,” Ann. Geophys., vol. 62,
“A geophysical multi-parametric
no. 1, 2019.
analysis of hydrothermal activity at
Dallol, Ethiopia,” J. African Earth Sci.,
[5] G. Currenti, C. del Negro, vol. 58, no. 5, pp. 812-819, Dec. 2010.
V. Lapenna, and L. Telesca,
“Multifractality in local geomagnetic [14] R. Carniel, “Characterization of
field at Etna volcano, Sicily (southern volcanic regimes and identification of
Italy),” Nat. Hazards Earth Syst. Sci., significant transitions using geophysical
vol. 5, no. 4, pp. 555-559, 2005. data: A review,” Bull. Volcanol., vol. 76,
no. 8, pp. 1-22, Jul. 2014.
[6] R. M. Green, M. S. Bebbington, S.
J. Cronin, and G. Jones, “Geochemical [15] M. Tárraga, J. Martí, R. Abella,
precursors for eruption repose length,” R. Carniel, and C. López, “Volcanic
Geophys. J. Int., vol. 193, no. 2, pp. 855- tremors: Good indicators of change
873, 2013. in plumbing systems during volcanic
eruptions,” J. Volcanol. Geotherm. Res.,
[7] S. Ebmeier et al., “Satellite geodesy vol. 273, pp. 33-40, 2014.
for volcano monitoring in the
Sentinel-1 and SAR constellation era,” [16] R. Carniel and M. Tárraga, “Can
in International Geoscience and Remote tectonic events change volcanic tremor
Sensing Symposium (IGARSS), 2019, pp. at Stromboli?,” Geophys. Res. Lett., vol.
5465-5467. 33, no. 20, 2006.

17
Updates in Volcanology – Transdisciplinary Nature of Volcano Science

[17] S. Dumont et al., “The dynamics slow-rupture earthquakes,” Nat. Geosci.,

of a long-lasting effusive eruption vol. 7, no. 1, pp. 71-75, 2014.
modulated by Earth tides,” Earth Planet.
Sci. Lett., 2020. [28] E. Marchetti and M. Ripepe,
“Stability of the seismic source during
[18] G. Tamburello et al., “Periodic effusive and explosive activity at
volcanic degassing behavior: The Mount Stromboli Volcano,” Geophys. Res. Lett.,
Etna example,” Geophys. Res. Lett., 2013. vol. 32, no. 3, pp. 1-5, 2005.

[19] V. M. Zobin, Introduction to Volcanic [29] R. Carniel, “Comments on the paper

Seismology: Third Edition. 2016. ‘Automatic detection and discrimination
of volcanic tremors and tectonic
[20] S. R. McNutt, “Volcano seismology earthquakes: An application to Ambrym
and monitoring for eruptions,” Int. volcano, Vanuatu’ by Daniel Rouland,
Handb. Earthq. Eng. Seismol., Denis Legrand, Mikhail Zhizhin and
pp. 383-406, 2002. Sylvie Vergniolle [J. Volcanol. Geotherm.
Res. 181,” J. Volcanol. Geotherm. Res., vol.
[21] B. Chouet, “Volcano Seismology,” 194, no. 1-3, pp. 61-62, Jul. 2010.
Pure Appl. Geophys., vol. 160, no. 3,
pp. 739-788, 2003. [30] A. Jolly, C. Caudron, T. Girona, B.
Christenson, and R. Carniel, “‘Silent’
[22] I. Molina, H. Kumagai, and H. Dome Emplacement into a Wet Volcano:
Yepes, “Resonances of a volcanic conduit Observations from an Effusive Eruption
triggered by repetitive injections of an at White Island (Whakaari), New
ash-laden gas,” Geophys. Res. Lett., 2004. Zealand in Late 2012,” Geosci., vol. 10,
no. 4, pp. 1-13, 2020.
[23] R. M. Iverson, “Dynamics of
seismogenic volcanic extrusion resisted [31] G. Cortés, R. Carniel, M.
by a solid surface plug, Mount St. A. Mendoza, and P. Lesage,
Helens, 2004-2005,” US Geol. Surv. Prof. “Standardization of Noisy
Pap., no. 1750, pp. 425-460, 2008. Volcanoseismic Waveforms as a Key Step
toward Station-Independent, Robust
[24] B. R. Julian, “Volcanic tremor: Automatic Recognition,” Seismol. Res.
nonlinear excitation by fluid flow,” J. Lett., vol. 90, no. 2 A, pp. 581-590, 2019.
Geophys. Res., vol. 99, no. B6, pp. 11,
811-859, 877, 1994. [32] A. Duque et al., “Exploring the
unsupervised classification of seismic
[25] M. Hellweg, “Physical models for events of Cotopaxi volcano,” J. Volcanol.
the source of Lascar’s harmonic tremor,” Geotherm. Res., p. 107009, 2020.
J. Volcanol. Geotherm. Res., vol. 101,
no. 1-2, pp. 183-198, 2000. [33] V. Janoušek, C. M. Farrow, and
V. Erban, “Interpretation of whole-
[26] J. B. Johnson, J. M. Lees, A. Gerst, rock geochemical data in igneous
D. Sahagian, and N. Varley, “Long- geochemistry: Introducing Geochemical
period earthquakes and co-eruptive Data Toolkit (GCDkit),” J. Petrol., vol.
dome inflation seen with particle image 47, no. 6, pp. 1255-1259, 2006.
velocimetry,” Nature, vol. 456, no. 7220,
pp. 377-381, 2008. [34] M. Petrelli and D. Perugini, “Solving
petrological problems through machine
[27] C. J. Bean, L. De Barros, I. Lokmer, learning: the study case of tectonic
J.-P. Métaxian, G. O’Brien, and S. discrimination using geochemical and
Murphy, “Long-period seismicity in the isotopic data,” Contrib. to Mineral.
shallow volcanic edifice formed from Petrol., 2016.

18
Machine Learning in Volcanology: A Review
DOI: [Link]

[35] M. S. M. Bolton et al., “Machine Analysis of volcanic tremor at

learning classifiers for attributing tephra Stromboli,” Nat. Hazards Earth Syst.
to source volcanoes: an evaluation of Sci., vol. 6, no. 6, pp. 903-909 ST-On the
methods for Alaska tephras,” J. Quat. singular values decoupling in, 2006.
Sci., 2020.
[45] A. Tharwat, “Principal component
[36] M. Brandmeier and G. Wörner, analysis - a tutorial,” Int. J. Appl. Pattern
“Compositional variations of ignimbrite Recognit., 2016.
magmas in the Central Andes over the
past 26 Ma — A multivariate statistical [46] J. Guo, H. Li, J. Ning, W. Han,
perspective,” Lithos, 2016. W. Zhang, and Z. S. Zhou, “Feature
dimension reduction using stacked
[37] K. Ueki, H. Hino, and T. Kuwatani, sparse auto-encoders for crop
“Geochemical discrimination and classification with multi-temporal,
characteristics of magmatic tectonic quad-pol SAR Data,” Remote Sens., 2020.
settings: A machine-learning-based
approach,” Geochemistry, Geophys. [47] A. Tharwat, T. Gaber, A. Ibrahim,
Geosystems, 2018. and A. E. Hassanien, “Linear
discriminant analysis: A detailed
[38] J. Hong, C. Gan, and J. Liu, tutorial,” AI Commun., 2017.
“Prediction of REEs in OIB by major
[48] S. Solorio-Fernández, J. A.
elements based on machine learning,”
Carrasco-Ochoa, and J. F. Martínez-
Earth Sci. Front., 2019.
Trinidad, “Ranking based unsupervised
feature selection methods: An empirical
[39] A. Burkov, The Hundred-Page
comparative study in high dimensional
Machine Learning Book. Andriy Burkov
datasets,” in Lecture Notes in Computer
(January 13, 2019), 2019.
Science (including subseries Lecture Notes
in Artificial Intelligence and Lecture Notes
[40] T. M. Mitchell, “The Discipline of
in Bioinformatics), 2018.
Machine Learning,” Mach. Learn., 2006.
[49] P. Caroline Cynthia and S. Thomas
[41] G. Cortés, M. Carmen Benitez, L.
George, “An Outlier Detection Approach
Garcia, I. Alvarez, and J. M. Ibanez, “A
on Credit Card Fraud Detection Using
Comparative Study of Dimensionality
Machine Learning: A Comparative
Reduction Algorithms Applied to
Analysis on Supervised and
Volcano-Seismic Signals,” IEEE J. Sel.
Unsupervised Learning,” in Intelligence
Top. Appl. Earth Obs. Remote Sens., 2016.
in Big Data Technologies---Beyond the
Hype, 2021, pp. 125-135.
[42] G. Cabras, R. Carniel, and J.
Wasserman, “Signal enhancement with [50] S. J. Pan and Q. Yang, “A survey on
generalized ICA applied to Mt. Etna transfer learning,” IEEE Trans. Knowl.
Volcano, Italy,” Boll. di Geofis. Teor. ed Data Eng., vol. 22, no. 10, pp. 1345-1359,
Appl., 2010. 2010.

[43] R. Carniel, G. Cabras, M. Ichihara, [51] M. Titos, A. Bueno, L. García, C.

and M. Takeo, “Filtering wind in Benítez, and J. C. Segura, “Classification
infrasound data by non-negative matrix of Isolated Volcano-Seismic Events
factorization,” Seismol. Res. Lett., vol. 85, Based on Inductive Transfer Learning,”
no. 5, pp. 1056-1062, 2014. IEEE Geosci. Remote Sens. Lett., vol. 17,
no. 5, pp. 869-873, 2020.
[44] R. Carniel, F. Barazza, M. Tárraga,
and R. Ortiz, “On the singular values [52] A. Bueno, C. Benitez, S. De Angelis,
decoupling in the Singular Spectrum A. Diaz Moreno, and J. M. Ibanez,

19
Updates in Volcanology – Transdisciplinary Nature of Volcano Science

“Volcano-Seismic Transfer Learning a new SOM approach,” J. Volcanol.

and Uncertainty Quantification with Geotherm. Res., 2013.
Bayesian Neural Networks,” IEEE Trans.
Geosci. Remote Sens., vol. 58, no. 2, [63] A. Köhler, M. Ohrnberger, and F.
pp. 892-902, 2020. Scherbaum, “Unsupervised pattern
recognition in continuous seismic
[53] G. Van Rossum and F. L. Drake, wavefield records using Self-Organizing
Python 3 Reference Manual. Scotts Valley, Maps,” Geophys. J. Int., 2010.
CA: CreateSpace, 2009.
[64] R. Carniel, “Neural networks
[54] R Core Team, “R: A Language and and dynamical system techniques
Environment for Statistical Computing.” for volcanic tremor analysis,” Ann. di
Vienna, Austria, 2013. Geofis., vol. 39, no. 2, pp. 241-252, 1996.

[55] K. Arnold, J. Gosling, and D. [65] A. M. Esposito, L. D’Auria, F.

Holmes, The Java programming language. Giudicepietro, T. Caputo, and M.
Addison Wesley Professional, 2005. martini, “Neural analysis of seismic
data: Applications to the monitoring of
[56] D. Flanagan, JavaScript: the Mt. Vesuvius,” Ann. Geophys., vol. 56,
definitive guide. “ O’Reilly Media, Inc.,” no. 4, 2013.
2006.
[66] S. Scarpetta et al., “Automatic
[57] J. Bezanson, A. Edelman, S. classification of seismic signals at Mt.
Karpinski, and V. B. Shah, “Julia: A fresh Vesuvius volcano, Italy, using neural
approach to numerical computing,” networks,” Bull. Seismol. Soc. Am., vol.
SIAM Rev., 2017. 95, no. 1, pp. 185-196, 2005.

[58] M. Odersky, L. Spoon, and B. [67] M. Masotti, S. Falsaperla, H. Langer,

Venners, Programming in scala. Artima S. Spampinato, and R. Campanini,
Inc, 2008. “Application of Support Vector Machine
to the classification of volcanic tremor at
[59] github, “GitHub.” 2020. Etna, Italy,” Geophys. Res. Lett., vol. 33,
no. 20, 2006.
[60] C. X. Ren, A. Peltier, V. Ferrazzini,
B. Rouet-Leduc, P. A. Johnson, and F. [68] M. Titos, A. Bueno, L. García, M. C.
Brenguier, “Machine Learning Reveals Benítez, and J. Ibañez, “Detection and
the Seismic Signature of Eruptive Classification of Continuous Volcano-
Behavior at Piton de la Fournaise Seismic Signals with Recurrent Neural
Volcano,” Geophys. Res. Lett., vol. 47, Networks,” IEEE Trans. Geosci. Remote
no. 3, p. e2019GL085523, 2020. Sens., vol. 57, no. 4, pp. 1936-1948, 2019.

[61] R. Carniel, L. Barbui, and A. D. [69] M. Beyreuther, R. Carniel, and J.

Jolly, “Detecting dynamical regimes by Wassermann, “Continuous Hidden
Self-Organizing Map (SOM) analysis: Markov Models: Application to
An example from the March 2006 automatic earthquake detection and
phreatic eruption at Raoul Island, New classification at Las Canãdas caldera,
Zealand Kermadec Arc,” Boll. di Geofis. Tenerife,” J. Volcanol. Geotherm. Res.,
Teor. ed Appl., 2013. 2008.

[62] R. Carniel, A. D. Jolly, and L. [70] M. Beyreuther and J. Wassermann,

Barbui, “Analysis of phreatic events at “Hidden semi-Markov Model based
Ruapehu volcano, New Zealand using earthquake classification system using

20
Machine Learning in Volcanology: A Review
DOI: [Link]

Weighted Finite-State Transducers,” [80] C. Cortes and V. Vapnik, “Support-

Nonlinear Process. Geophys., vol. 18, no. 1, vector networks,” Mach. Learn., vol. 20,
pp. 81-89, 2011. no. 3, pp. 273-297, 1995.

[71] G. Cortés et al., “Evaluating [81] B. Krishnapuram, L. Carin, M. A.

robustness of a HMM-based T. Figueiredo, and A. J. Hartemink,
classification system of volcano- “Sparse multinomial logistic regression:
seismic events at COLIMA and Fast algorithms and generalization
Popocatepetl volcanoes,” in International bounds,” IEEE Trans. Pattern Anal.
Geoscience and Remote Sensing Mach. Intell., 2005.
Symposium (IGARSS), 2009, vol. 2, pp.
II1012-II1015. [82] G. Cabras, R. Carniel, and J. Jones,
“Non-negative Matrix Factorization:
[72] G. Cortés, L. García, I. Álvarez, C. An application to Erta ‘Ale volcano,
Benítez, Á. de la Torre, and J. Ibáñez, Ethiopia,” Boll. di Geofis. Teor. ed Appl.,
“Parallel System Architecture (PSA): vol. 53, no. 2, pp. 231-242, 2012.
An efficient approach for automatic
recognition of volcano-seismic events,” [83] B. Krishnapuram, L. Carin, and A. J.
J. Volcanol. Geotherm. Res., 2014. Hartemink, “Joint classifier and feature
optimization for cancer diagnosis using
[73] F. Pedregosa et al., “Scikit-learn: gene expression data,” in Proceedings
Machine Learning in {P}ython,” J. of the Annual International Conference
Mach. Learn. Res., vol. 12, pp. 2825- on Computational Molecular Biology,
2830, 2011. RECOMB, 2003.

[74] M. Charrad, N. Ghazzali, V. Boiteau, [84] B. T. Pham et al., “Ensemble

and A. Niknafs, “Nbclust: An R package modeling of landslide susceptibility
for determining the relevant number using random subspace learner and
of clusters in a data set,” J. Stat. Softw., different decision tree classifiers,”
2014. Geocarto Int., 2020.

[85] L. R. Rabiner and R. W. Schafer,

[75] T. Kohonen, “Self-organized
“Introduction to digital speech
formation of topologically correct
processing,” Found. Trends Signal
feature maps,” Biol. Cybern., 1982.
Process., 2007.
[76] T. Kohonen, Self-organizing maps, [86] P. Alasonati, J. Wassermann, and
3rd ed. Berlin: Springer, 2001. M. Ohrnberger, “Signal classification by
wavelet-based hidden Markov models:
[77] S. Hochreiter and J. Schmidhuber, application to seismic signals of volcanic
“Long Short-Term Memory,” Neural origin,” in Statistics in Volcanology, 2018.
Comput., vol. 9, no. 8, pp. 1735-1780,
1997. [87] A. Boué, P. Lesage, G. Cortés, B.
Valette, and G. Reyes-Dávila, “Real-time
[78] P. McCullagh and J. A. Nelder, eruption forecasting using the material
Generalized Linear Models, Second Failure Forecast Method with a Bayesian
Edition (Monographs on Statistics and approach,” J. Geophys. Res. Solid Earth,
Applied Probability). 1989. vol. 120, no. 4, pp. 2143-2161, 2015.

[79] T. Hastie, R. Tibshirani, and J. [88] M. Tárraga, R. Carniel, R. Ortiz,

Friedman, Elements of Statistical and A. García, “Chapter 13 The
Learning 2nd ed. 2009. Failure Forecast Method: Review and

21
Updates in Volcanology – Transdisciplinary Nature of Volcano Science

Application for the Real-Time Detection and analysis of repeating earthquakes

of Precursory Patterns at Reawakening Application to Mount St. Helens and
Volcanoes,” Dev. Volcanol., vol. 10, no. C, Redoubt volcanoes,” in Seismological
pp. 447-469, 2008. Society of America Annual Meeting,
2016.
[89] J. P. Jones, R. Carniel, and S. D.
Malone, “Subband decomposition and [96] A. D. Jolly et al., “Seismo-acoustic
reconstruction of continuous volcanic evidence for an avalanche driven
tremor,” J. Volcanol. Geotherm. Res., vol. phreatic eruption through a beheaded
213-214, pp. 98-115, 2012. hydrothermal system: An example
from the 2012 Tongariro eruption,” J.
[90] H. Langer, S. Falsaperla, T. Volcanol. Geotherm. Res., vol. 286, pp.
Powell, and G. Thompson, “Automatic 331-347, 2014.
classification and a-posteriori analysis
of seismic event identification at [97] A. Messina and H. Langer, “Pattern
Soufrière Hills volcano, Montserrat,” J. recognition of volcanic tremor data on
Volcanol. Geotherm. Res., vol. 153, no. 1-2 Mt. Etna (Italy) with KKAnalysis-A
SPEC. ISS., pp. 1-10, 2006. software program for unsupervised
classification,” Comput. Geosci., vol. 37,
[91] J. M. Ibáñez, C. Benítez, L. A. no. 7, pp. 953-961, 2011.
Gutiérrez, G. Cortés, A. García-Yeguas,
and G. Alguacil, “The classification [98] S. Falsaperla, S. Graziani,
of seismo-volcanic signals using G. Nunnari, and S. Spampinato,
Hidden Markov Models as applied to “Automatic classification of volcanic
the Stromboli and Etna volcanoes,” J. earthquakes by using multi-layered
Volcanol. Geotherm. Res., 2009. neural networks,” Nat. Hazards, 1996.

[92] D. N. Green and J. Neuberg, [99] H. Langer, S. Falsaperla, M. Masotti,

“Waveform classification of volcanic R. Campanini, S. Spampinato, and A.
low-frequency earthquake swarms Messina, “Synopsis of supervised and
and its implication at Soufrière Hills unsupervised pattern classification
Volcano, Montserrat,” J. Volcanol. techniques applied to volcanic tremor
Geotherm. Res., vol. 153, no. 1-2 SPEC. data at Mt Etna, Italy,” Geophys. J. Int.,
ISS., pp. 51-63, 2006. 2009.

[93] R. Villegas, R. Carniel, I. Petrinovic, [100] M. Titos, A. Bueno, L. Garcia, and

and C. Balbis, “Clusters of long-period C. Benitez, “A Deep Neural Networks
(LP) seismic events at the Irazú Volcano: Approach to Automatic Recognition
what are they telling us?,” J. South Am. Systems for Volcano-Seismic Events,”
Earth Sci., no. under final revision, IEEE J. Sel. Top. Appl. Earth Obs. Remote
2020. Sens., 2018.

[94] W. I. Sevilla, L. A. Jumawan, [101] G. Curilem, J. Vergara, G.

C. J. Clarito, M. A. Quintia, A. A. Fuentealba, G. Acuña, and M. Chacón,
Dominguiano, and R. U. Solidum, “Classification of seismic signals at
“Improved 1D velocity model and deep Villarrica volcano (Chile) using neural
long-period earthquakes in Kanlaon networks and genetic algorithms,” J.
Volcano, Philippines: Implications Volcanol. Geotherm. Res., vol. 180, no. 1,
for its magmatic system,” J. Volcanol. pp. 1-8, 2009.
Geotherm. Res., 2020.
[102] J. P. Canário et al., “In-depth
[95] A. J. Hotovec-Ellis and C. Jeffries, comparison of deep artificial neural
“Near real-time detection, clustering, network architectures on seismic events

22
Machine Learning in Volcanology: A Review
DOI: [Link]

classification,” J. Volcanol. Geotherm. [110] M. C. Benítez et al., “Continuous

Res., vol. 401, 2020. HMM-based seismic-event classification
at deception Island, Antarctica,” in IEEE
[103] B. Pradhan and S. Lee, “Landslide Transactions on Geoscience and Remote
susceptibility assessment and factor Sensing, 2007.
effect analysis: backpropagation
artificial neural networks and their [111] M. Bicego, C. Acosta-Munoz, and
comparison with frequency ratio and M. Orozco-Alzate, “Classification of
bivariate logistic regression modelling,” seismic volcanic signals using hidden-
Environ. Model. Softw., 2010. markov-model-based generative
embeddings,” IEEE Trans. Geosci. Remote
[104] G. F. Manley et al., “Understanding Sens., vol. 51, no. 6, pp. 3400-3409,
the timing of eruption end using 2013.
a machine learning approach to
classification of seismic time series,” J. [112] P. B. Dawson, M. C. Benítez,
Volcanol. Geotherm. Res., 2020. B. A. Chouet, D. Wilson, and P. G.
Okubo, “Monitoring very-long-period
[105] M. Curilem et al., “Pattern seismicity at Kilauea Volcano, Hawaii,”
recognition applied to seismic signals of Geophys. Res. Lett., 2010.
Llaima volcano (Chile): An evaluation
of station-dependent classifiers,” J. [113] N. Trujillo-Castrillón, C. M.
Volcanol. Geotherm. Res., vol. 315, pp. Valdés-González, R. Arámbula-
15-27, 2016. Mendoza, and C. C. Santacoloma-
Salguero, “Initial processing of volcanic
[106] M. Malfante, M. Dalla Mura, J. seismic signals using Hidden Markov
P. Metaxian, J. I. Mars, O. Macedo, Models: Nevado del Huila, Colombia,” J.
and A. Inza, “Machine Learning for Volcanol. Geotherm. Res., 2018.
Volcano-Seismic Signals: Challenges and
Perspectives,” IEEE Signal Process. Mag., [114] C. cassisi, M. Prestifilippo, A.
2018. Cannata, P. Montalto, D. Patanè, and
E. Privitera, “Probabilistic Reasoning
[107] I. M. Murwantara, P. Yugopuspito, Over Seismic Time Series: Volcano
and R. Hermawan, “Comparison of Monitoring by Hidden Markov Models
machine learning performance for at Mt. Etna,” Pure Appl. Geophys., 2016.
earthquake prediction in Indonesia
using 30 years historical data,” [115] A. Maggi, V. Ferrazzini, C.
Telkomnika (Telecommunication Comput. Hibert, F. Beauducel, P. Boissier, and
Electron. Control., 2020. A. Amemoutou, “Implementation of a
multistation approach for automated
[108] C. Hibert, F. Provost, J. P. Malet, event classification at Piton de la
A. Maggi, A. Stumpf, and V. Ferrazzini, Fournaise volcano,” Seismol. Res. Lett.,
“Automatic identification of rockfalls 2017.
and volcano-tectonic earthquakes at the
Piton de la Fournaise volcano using a [116] P. E. E. Lara et al., “Automatic
Random Forest algorithm,” J. Volcanol. multichannel volcano-seismic
Geotherm. Res., 2017. classification using machine learning
and EMD,” IEEE J. Sel. Top. Appl. Earth
[109] D. E. Dempsey, S. J. Cronin, S. Mei, Obs. Remote Sens., 2020.
and A. W. Kempa-Liehr, “Automatic
precursor recognition and real-time [117] M. Beyreuther, R. Barsch, L.
forecasting of sudden explosive volcanic Krischer, T. Megies, Y. Behr, and J.
eruptions at Whakaari, New Zealand,” Wassermann, “ObsPy: A python toolbox
Nat. Commun., 2020. for seismology,” Seismol. Res. Lett., 2010.

23
Updates in Volcanology – Transdisciplinary Nature of Volcano Science

[118] T. Lecocq, C. Caudron, and F. [127] D. Cervelli, P. Cervelli, T. Parker,

Brenguier, “Msnoise, a python package and T. Murray, “SWARM Seismic Wave
for monitoring seismic velocity changes Analysis and Real-time Monitor: User
using ambient seismic noise,” Seismol. Manual and Reference Guide,” 2020.
Res. Lett., 2014. [Online]. Available: [Link]
[Link]/software/swarm/[Link].
[119] A. Bueno et al., “PICOSS: Python
Interface for the Classification of [128] J. A. Cortés, J. L. Palma, and M.
Seismic Signals,” Comput. Geosci., Wilson, “Deciphering magma mixing:
2020. The application of cluster analysis
to the mineral chemistry of crystal
[120] H. Buurman and M. E. West, populations,” J. Volcanol. Geotherm. Res.,
“Seismic precursors to volcanic 2007.
explosions during the 2006 eruption of
Augustine Volcano,” US Geol. Surv. Prof. [129] U. Morgenstern, C. J. Daughney,
Pap., 2010. G. Leonard, D. Gordon, F. M. Donath,
and R. Reeves, “Using groundwater
[121] G. Cortés, R. Carniel, P. Lesage, age and hydrochemistry to understand
and M. A. Mendoza, “VULCAN. sources and dynamics of nutrient
ears: Volcano-seismic Unsupervised contamination through the catchment
Labelling and ClAssificatioN into Lake Rotorua, New Zealand,”
Embedded in A Real-time Scenario,” Hydrol. Earth Syst. Sci., 2015.
2020. [Online]. Available: https://
[Link]/project/id/749249/it. [130] M. O. Awaleh et al., “Geochemical,
[Accessed: 20-Aug-2020]. multi-isotopic studies and geothermal
potential evaluation of the complex
[122] G. Cortés, R. Carniel, P. Lesage, Djibouti volcanic aquifer (republic of
and M. A. Mendoza, “pyVERSO - Djibouti),” Appl. Geochemistry, 2018.
software for building and evaluating
Volcano-Seismic Recognition (VSR) [131] F. Barette, S. Poppe, B. Smets, M.
system,” 2020. . Benbakkar, and M. Kervyn, “Spatial
variation of volcanic rock geochemistry
[123] E. D. Cambridge University, in the Virunga Volcanic Province:
“HTK - Hidden Markov Model Toolkit,” Statistical analysis of an integrated
2020. [Online]. Available: [Link] database,” J. African Earth Sci., 2017.
[Link]/.
[132] The Mathworks Inc., “MATLAB -
[124] G. Cortés, R. Carniel, P. Lesage, MathWorks,” [Link]/
and M. A. Mendoza, “geoStudio & products/matlab, 2020. .
liveVSR software,” 2020. .
[133] E. S. Schandl and M. P. Gorton,
[125] G. Cortés, R. Carniel, M. “Application of high field strength
A. Mendoza, and P. Lesage, elements to discriminate tectonic
“VSR Databases used in article settings in VMS environments,” Econ.
‘Standardization of noisy volcano- Geol., 2002.
seismic waveforms as a key step towards
station-independent, robust automatic [134] J. A. Pearce and J. R. Cann,
recognition.’” Zenodo, 2018. “Tectonic setting of basic volcanic
rocks determined using trace element
[126] P. Lesage, “Interactive Matlab analyses,” Earth Planet. Sci. Lett., 1973.
software for the analysis of seismic
volcanic signals,” Comput. Geosci., vol. [135] C. Li, N. T. Arndt, Q. Tang,
35, no. 10, pp. 2137-2144, 2009. and E. M. Ripley, “Trace element

24
Machine Learning in Volcanology: A Review
DOI: [Link]

indiscrimination diagrams,” Lithos. volcanic areas: New age constraints for

2015. the Pleistocene magmatism of central
Italy,” Quat. Geochronol., 2017.
[136] C. A. Snow, “A reevaluation of
tectonic discrimination diagrams and a [144] A. M. Esposito, G. Alaia, F.
new probabilistic approach using large Giudicepietro, L. Pappalardo, and M.
geochemical databases: Moving beyond D’Antonio, “Unsupervised Geochemical
binary and ternary plots,” J. Geophys. Analysis of the Eruptive Products of
Res. Solid Earth, 2006. Ischia, Vesuvius and Campi Flegrei,”
in Smart Innovation, Systems and
[137] S. P. Verma and J. S. Armstrong- Technologies, 2021.
Altrin, “New multi-dimensional
diagrams for tectonic discrimination [145] Y. Zhao, Y. Zhang, M. Geng, J.
of siliciclastic sediments and their Jiang, and X. Zou, “Involvement of
application to Precambrian basins,” Slab-Derived Fluid in the Generation
Chem. Geol., 2013. of Cenozoic Basalts in Northeast China
Inferred From Machine Learning,”
[138] S. P. Verma and L. Díaz-González, Geophys. Res. Lett., 2019.
“New discriminant-function-based
multidimensional discrimination of [146] B. C. Dye and G. Morra, “Machine
mid-ocean ridge and oceanic plateau,” learning as a detection method of
Geosci. Front., 2020. Strombolian eruptions in infrared
images from Mount Erebus, Antarctica,”
[139] CoDaPack, “CoDaPack - Phys. Earth Planet. Inter., 2020.
Compositional Data Package,” 2020.
[Online]. Available: [Link] [147] G. Moreno Chávez, J. Villa, D.
[Link]/[Link]. Sarocchi, and E. González-Ramírez,
[Accessed: 20-Aug-2020]. “A method and software solution for
classifying clast roundness based on
[140] C. H. Weiß, “StatSoft, Inc., Tulsa, the radon transform,” Comput. Geosci.,
OK.: STATISTICA, Version 8,” AStA 2020.
Adv. Stat. Anal., 2007.
[148] D. Sinitò et al., “I-PETER
[141] A. B. Beaudoin and R. H. King, (Interactive platform to experience
“Using discriminant function analysis tours and education on the rocks): A
to identify Holocene tephras based on virtual system for the understanding
magnetite composition: a case study and dissemination of mineralogical-
from the Sunwapta Pass area, Jasper petrographic science,” Pattern Recognit.
National Park.,” Can. J. Earth Sci., 1986. Lett., 2020.

[142] A. J. Bourne et al., “Distal tephra [149] Y. Sunaga, R. Natsuaki, and A.

record for the last ca 105,000 years Hirose, “Proposal of complex-valued
from core PRAD 1-2 in the central convolutional neural networks for
Adriatic Sea: Implications for marine similar land-shape discovery in
tephrostratigraphy,” Quat. Sci. Rev., interferometric synthetic aperture
2010. radar,” in Lecture Notes in Computer
Science (including subseries Lecture Notes
[143] M. Petrelli, R. Bizzarri, D. in Artificial Intelligence and Lecture Notes
Morgavi, A. Baldanza, and D. in Bioinformatics), 2018.
Perugini, “Combining machine
learning techniques, microanalyses [150] D. Phiri, M. Simwanda, S. Salekin,
and large geochemical datasets for V. R. Nyirenda, Y. Murayama, and M.
tephrochronological studies in complex Ranagalage, “Sentinel-2 Data for Land

25
Updates in Volcanology – Transdisciplinary Nature of Volcano Science

Cover/Use Mapping: A Review,” Remote Imaging,” in Lecture Notes in Computer

Sens., vol. 12, no. 14, p. 2291, 2020. Science (including subseries Lecture Notes
in Artificial Intelligence and Lecture Notes
[151] N. Anantrasirichai, J. Biggs, F. in Bioinformatics), 2018.
Albino, P. Hill, and D. Bull, “Application
of Machine Learning to Classification [159] M. C. Burl et al., “Learning to
of Volcanic Deformation in Routinely recognize volcanoes on Venus,” Mach.
Generated InSAR Data,” J. Geophys. Res. Learn., 1998.
Solid Earth, vol. 123, no. 8, pp. 6592-
6606, 2018. [160] T. F. Stepinski, S. Ghosh, and R.
Vilalta, “Machine learning for automatic
[152] A. Krizhevsky, I. Sutskever, and G. mapping of planetary surfaces,” in
E. Hinton, “ImageNet classification with Proceedings of the National Conference on
deep convolutional neural networks,” in Artificial Intelligence, 2007.
Advances in Neural Information Processing
Systems, 2012, vol. 2, pp. 1097-1105.

[153] S. Valade et al., “Towards global

volcano monitoring using multisensor
sentinel missions and artificial
intelligence: The MOUNTS monitoring
system,” Remote Sens., 2019.

[154] A. Si et al., “Debris flow

susceptibility assessment using the
integrated random forest based steady-
state infinite slope method: A case study
in Changbai Mountain, China,” Water
(Switzerland), 2020.

[155] L. Li, C. Solana, F. Canters, and

M. Kervyn, “Testing random forest
classification for identifying lava flows
and mapping age groups on a single
Landsat 8 image,” J. Volcanol. Geotherm.
Res., 2017.

[156] D. T. Bui, P. Tsangaratos, V.

T. Nguyen, N. Van Liem, and P. T.
Trinh, “Comparing the prediction
performance of a Deep Learning Neural
Network model with conventional
machine learning models in landslide
susceptibility assessment,” Catena,
2020.

[157] D. Carbone, D. Gibert, J. Marteau,

M. Diament, L. Zuccarello, and E.
Galichet, “An experiment of muon
radiography at Mt Etna (Italy),” Geophys.
J. Int., 2013.

[158] G. Yang, D. Ireland, R. Kaiser, and

D. Mahon, “Machine Learning for Muon

26