Despite Costa Rica’s small
size (a mere 
), it is the site of four converging tectonic
plates — the Cocos, Caribbean, Nazca, and South American plates — as well as a
small block, referred to as the Panama Block. OVSICORI comprises approximately 
geodynamic GPS stations deployed throughout the
country. These stations, responsible for recording Costa Rica’s seismic
activity, registered a mean of 
earthquakes per year (see figure 1).
Fig. 1. (a) Tectonic plates convergence in Costa Rican territory. (b)
Visualization of OVSICORI monitoring stations visualized by the Costa Rican
Institute of Technology
Earthquakes are
conventionally documented with two types of points, the hypocenters, which are
the point at which the shock is generated, with longitude, latitude and depth
and the epicenters, which are the projection of the hypocenters into the
surface of the earth. Plinius visualizes the earthquake hypocenters that
occurred in Costa Rica in 32 years (
).
We agree with Li, Y. et al. [1] in that the task of analysis should be based on continuous
process feedback from domain experts. According to seismologists, earthquakes
can result from three causes: (1) local faults, (2) volcanic activity, and (3)
tectonic plate subduction. Earthquakes are additionally classified by their magnitude
on the Richter scale.
Earthquakes caused by local
faults or volcanic activity are generally superficial while events caused by
tectonic faults have typically deeper origins. We have defined a point sets 
such that 
where 
is the earthquakes’ epicenter and 
its depth. In other words, 
is the earthquakes’ hypocenter.
With the collaboration of
seismologists, we also define that if 
the earthquakes are associated with local faults or
volcanic activity, and if 
these are produced by tectonic plates subduction.
Another point set is defined
by such that
where is the earthquakes’ epicenter and its magnitude on the Richter scale. For this set, we
established three intensity categories:
•
low intensity;
•
medium intensity;
•
hight intensity, Costa
Rica does not have the capacity for generating earthquakes with magnitude upper
8.
Based on this classification, we made a study
of the data. Some of the more general results are presented in figure 2.
According to graphic 1 and 2 (figure 2), the
most of earthquakes are produced by local faults and with low intensity. In
this period the earthquakes distribution by year is very irregular,
highlighting 
like a year with more
earthquakes (
) and 
with less number
of earthquakes (
).
Fig. 2. Graphic 1: Distribution of the
number of earthquakes per year based on the cause. Graphic 2: Distribution of
the number of earthquakes per year based on the intensity. Graphic 3:
Distribution of the number of earthquakes per months. Graphic 4: Distribution
of the most significant earthquakes per year
From graphic 3, we can deduce that no month
has significantly higher seismic activity than the others. The last graphic (4)
shows that earthquakes with greater magnitude occurred each year are produced
due to local faults and clearly indicates the earthquake that occurred on
September 5, 
, one of the most
documented earthquakes in history, about which we made a visualization [2].
The sorts of everyday questions seismologists
must answer include: what kind of earthquakes are they,
according to their depth? How many earthquakes are produced by local faults? How many
earthquakes are generated by tectonic plate subduction? Where are there new
seismic swarms? How should seismic profiles be defined to determine high-risk
areas? What is the behavior of the tectonic plate and how is its geometry?
In order to answer these questions, most of
the data analysis today is carried out using two-dimensional systems involving
latitude and longitude and using different visual strategies to render depth
and magnitude, as shown in figure 3 and 4.
Fig. 3. Visualization of the magnitudes and
depths of the earthquakes that occurred during the year 2012 (OVSICORI)
Attributes like magnitude and
depth are depicted in 2D through color, size, and shape (see figure 3). For
example, Kyriakopoulos et al. and Meyer and Wischgoll [3] [4] use size and color to represent the
intensity of earthquakes. Dzwinel et al. [5] show the magnitude of the
earthquakes and their depth by different circle radiuses and colors; the same
scheme is used by Fairchild et al., but in 3D [6].
The cognitive burden of the aforementioned
image is extremely high and becomes even more complex when monitoring longer
time periods.
With this type of analysis, answering
questions such as which earthquakes are local or which are of tectonic origin,
which in turn depends on their depth, requires a detailed observation of the
symbols with which these variables are represented.
In addition, scientists use seismic profiles,
such as those shown in figure 4, to work with these questions, which itself
involves analyzing cross-section cuts based on depth.
Fig. 4. Seismicity and seismic profiles in
Costa Rica (OVSICORI)
The objective of profiling is to understand
plate movement in the Earth’s interior and its geometry. With these cross
sections, the seismologists isolate each part of the subduing zone to get an
idea of the geometry of the plate that is introduced underneath the other, more
superficial plate. This kind of analysis is very complicated and requires a
much higher capacity for abstraction that would be necessary if the
representation were in three dimensions and observable at any desired angle.
Plinius (named after the famous scientist of
the Roman Empire, Gaius Plinius Secundus who died after Mount Vesuvius erupted
in 79 AD) provides seismologists with a 3D tool to visually analyze earthquake
hypocenters, giving several filters for the data analysis and different options
to save the dataset examined.
First of all, Plinius was built on the needs
of the scientists of the domain (seismologists, volcanologists, and
geologists). Therefore, its design and development incorporated some elements
of both the theory of perception [7] and cognitive theory [8]. Elements of
interaction taxonomy [9] — filtering, selecting, abstract/elaborate,
overview/explore, and connect/relate — were integrated as well.
As Plinius was developed in a 3D environment,
we used the basic techniques proposed by Fernandez and Fetais [10] to enhance
the visual analysis.
Many systems include features for depicting
attributes such as intensity and depth. For example, EQviz [11], a
visualization tool for monitoring world earthquakes that only works in 2D, uses
colors ranging from yellow to red to represent earthquake magnitude, yellow
being the lowest and red the highest.
SeismoGIS [12] is a toolset that supports the
analysis and visualization of earthquakes measured by a seismometer station
network. However, its primary task is to locate and name the events. It is a
tool specifically for managing and monitoring the seismic events and is not
considered to meet any of scientists’ other essential requirements.
Wolfe et al. [13] present a visualization
system that examines seismic data using a volumetric scheme; the resulting 3D
images reveal the structure of the geological layers. Meanwhile, Patel et al.
[14] have developed a 2D toolset to analyze volumetric reflection data.
None of these visualization systems can
customize the analysis. That is to say, these systems do not have filters that
can customize the hypocenters that are being seen, either in depth, magnitude
or date. Much less is it possible with these tools to generate groups of data
either associated with defined points such as volcanoes or local faults, or
groups of the data related with profiles proposed at any angle, all of which
are essential tasks in the work of seismologists.
A system with interesting interactivity is
presented by Leonard et al. [15], who integrate some seismological
characteristics in a GIS system which allows multiple views of the relevant
information — albeit, being in 2D, with a significant cognitive burden. On the
other hand, the simulation of the 1906 San Francisco earthquake by Chourasia et
al. [16] uses a coloring scheme to reflect the intensity of shaking. Providing
visual clues to the audience about the duration of shaking and a major realism
to the ground motions, this is an engaging new approach to showing another kind
of qualitative data.
Other works [17, 18, 19, 20, 21, 22] are
related to systems that analyze seismic phenomena. Nevertheless, their focus is
the simulation, and most of them utilize volume-rendering methods to visualize
what kind of movement occurs.
For example, in the case of the work of Yuen
et al. [23], they developed a web service, WEB-IS, for remote visualization of
clustered seismic data within a GRID framework. Meyer & Wischgoll [4]
provide a simulation of the ground motion, both taking into account the
location of epicenter magnitudes and including, interestingly, fault lines.
Their simulation was performed through of finite element method (FEM) based on
time-varying tetrahedral meshes.
More similar to Plinius in intent are EQviz
[11], TerraVis [24], and a system developed by NEIC [25], all of which
visualize the hypocenters beneath the Earth’s surface. However, these systems
do not facilitate the resolution of many of the necessary tasks proposed and
needed by seismologists. In other words, these tools lack interactivity and
therefore it is not possible for scientists to use them to analyze specific
situations that are used in their daily work.
For these reasons, compared with all the
previous literature our work differs in several respects. Plinius takes
historical data and represents them in a 3D environment beneath the Earth’s
surface — in this case, beneath Costa Rica — allowing users to analyze the
tectonic dynamics of the country. It permits filtering for different variables —
such as date, magnitude, and depth — in any combination. Also, it allows users
to control two kinds of seismic profiles. With the first one, they can get the
cross section to visualize the plates’ tectonic structure. With the second one,
they can analyze seismic events around a particular point, such as a volcano.
Most of the analysis can be saved into a file with a specific set of data or
specific images of any cross-section or view.
It is commonly recognized that 3D data should
be displayed in 3D visualizations and 2D data in 2D visualizations. This
well-known affirmation is almost always right, but never righter than in the
case of this kind of data, because most research on hypocenters attempts to
show 3D information in 2D contexts.
Hypocenters' distributions are generated by
three factors: (1) the 3D geometry of tectonic plates, (2) local faults, and
(3) volcanic behaviors.
Because of this, our major design goal was to
provide a 3D visualization with a navigation system adapted to the needs of
scientists, and with unique tools specially designed for these needs. This
general visual paradigm was chosen based on a “target question methodology”
[14], which focuses on responding to the need (or question) to be resolved by
the visualization.
For 3D navigation, we use our
customized approach to the conventional Virtual Trackball,
Virtual Sphere or Sphere View [26].
The interface includes various dashboards: a
hypocenter dashboard, geographic dashboard, and a general data dashboard. Most of
these elements follow the trend called “embedded interaction”, as defined by
Saket et al. [27].
The hypocenter dashboard is
Plinius’ central dashboard and allows the user to filter data according to
depth, time and magnitude. Users can visualize all of the hypocenters codified
chromatically based on the magnitude, and it allows users to establish seismic
profiles based on their criteria.
The geographic dashboard
provides a visualization of different resolutions of the country’s geographic.
The "general data dashboard" presents information as a current number
of frames per second and allows users to visualize the location of GPS stations
and other places of reference (see figure 5).
Fig. 5. Plinius Dashboards
Various work sessions were held with OVSICORI
seismologists to assess Plinius and define the principal tasks required by the
tool.
As stated, seismologists often work with
questions such as: How many earthquakes are produced by faults? How many
earthquakes are generated by tectonic plate subduction? Where are these new
seismic swarms? How should seismic profiles be defined to determine high-risk
areas? What is the behavior of the tectonic plate and its geometric direction?
In order to answer questions like “How many
earthquakes are produced by faults?” or “How many earthquakes are generated by
tectonic plate subduction?” earthquakes are classified according to depth. For
instance, earthquakes with a depth between 35 and 200 km frequently result from
tectonic plate interaction. Earthquakes with a depth of less than 35 km are,
most of the time, caused by local faults and volcanic behaviors. Because of
this, a good visualization of deeper earthquakes would entail a good image of
plate surfaces sliding against each other. However, traditionally, graphic
strategies such as crosses and squares are used on the map to define depths and
magnitude (as shown in figure 3 and 4).
Plinius's hypocenter dashboard allows the user
to filter data according to depth, period, and magnitude. Users can visualize
(in 3D, allowing for different points of view and zoom possibilities) all of
the hypocenters at each range of depth, codified chromatically based on the
magnitude and period, using three double sliders to specify desired ranges.
So, if the user wishes to visualize an
earthquake arising from local faults or tectonic plates, the depth filter can
be used for this purpose. That’s especially useful to
work with questions like: “Where are there new seismic swarms?”, “How should
seismic profiles be defined to determine high-risk areas?”
Furthermore, there is a slider to control the
opacity of points, which is very useful because surface earthquakes are
significantly more frequent (80%) than deep earthquakes and therefore changing
the opacity of a case to case allows work with occlusion between data (see figure
6 and 7).
Fig. 6. Visualization of 
earthquakes less than 35 km deep, showing local fault areas
Fig. 7. Visualization of 
earthquakes more than 35 km deep, showing tectonic fault areas
Magnitude is another important parameter for
classifying earthquakes. Seismologists must be able to visualize those areas of
the country where high magnitude earthquakes have occurred and contrast them
with research data. In order to visualize earthquakes according to their
magnitude on Plinius, the third section of the dashboard shows a double slider
that can be used to adjust the range of magnitude.
Magnitude can also be visualized in different
colors using a chromatic code (figure 8). Green represents lower magnitude
earthquakes and red higher magnitude earthquakes. In addition, earthquakes with
more significant magnitudes (less than 20% of the total) are represented larger
than all of those with small magnitudes; this feature, together with the
“opacity slider”, allows scientists to fine-tune what they see.
Fig. 8. Example of earthquake
visualization where point color (green, yellow, red) and size are used to
represent earthquake magnitude
It is worth noting that
3D-perspective views always imply a focus-plus-context effect because in
natural human view the perspective makes faraway objects smaller and closer
objects bigger for the observer. This process is one of the advantages of
having a 3D view of more than 
hypocenters:
the seismologist can turn around, get closer, or move further away in order to
catch precisely the object of interest.
Perhaps the most important and time-consuming
task performed by seismologists is defining seismic profiles. The objective is
to determine vertical cross-section cuts and thereby analyze tectonic plate
subduction trends. As reflected in figure 4, this step is viewed in 2D.
Questions like, “what is the behavior of the tectonic
plate and its geometric direction?” and “how is the geometry of subduction
area?” are the kind of topics that the scientists answer with these profiles.
With Plinius, when the user indicates that
he/she wants to define a new profile, the tool moves the point of view to a
unique orthogonal place to facilitate the conventional method that scientists
use to introduce three points. The first two points define a line, and the
third establishes the width of the profile. In this way, a procedure that previously
took days now can be carried out in seconds (figure 9).
This feature of Plinius enables the scientist
to make several profiles in minutes, which is important because of data
occlusion. Usually, if you look at a cross-section of the hypocenter data, you
will see all the points behind the zone of interest. With Plinius, you can, in
seconds, cut out all these points of the perspective and concentrate in the
desired profile (typically used to see seismic subduction plates), testing over
and again if you do not get the result that you want. More details of work with seismic profiles in Plinius are described
in [28]
Fig.
9. Defining seismic profiles by Plinius
With this data, scientists can define the
shape of the tectonic plates and the geometry in which one is subduced into the
other. To observe this geometry in three dimensions, Plinius can visualize this
geometry and in turn mix it with all other data and filters, giving users the
ability to see this geometry in context (see figure 10).
Fig. 10. Geometry of the subduction zone
The objective of this filter is to determine
cylindrical cross-section cuts and thereby analyze local failures and volcanic
behaviors. This feature was asked for by the scientists when they were able to
see, for the first time, only surface earthquakes, they decided that it would
be useful to be able to isolate specific areas. These areas with points of
interest on the surface (e.g., volcanoes, local faults, cities, and coastal
locations), would be separated and could be analyzed quickly with close-ups and
rotations of the associated swarm of earthquakes (as shown in figure 6 above).
To do that, we use two points: the first is
the central position of the cylinder (usually a volcano, a local fault, or a
city of interest); the second is the ratio of the desired area, the width of
the cylinder (see figure 11). In this way, seismologists can isolate a local
area with all hypocenters underneath it, then potentially filter for
magnitudes, periods, or depths with the other sliders, as shown in figure 11
(Irazú Volcano in Costa Rica).
Fig. 11. Cylindrical area of interest
defined by seismologists, allowing visualization of the Irazú volcano
zone in Costa Rica
The
entire Plinius system was developed using Processing 3.0 (https://processing.org), which is a free software
platform developed by the Massachusetts Institute of Technology and
is functional on Windows, macOS, and Linux operating systems.
In other words, from Processing it is possible to export a tool
that works in any of these operating systems.
The
Processing platform is a Java dialect and uses as a 3D engine a version of
OpenGL, specifically OpenGLES. This configuration allows it to represent
moderately complex three-dimensional models with good features, in our
case the efficiency is good up to a maximum of one million points.
This system was tested in its efficiency with
two other development environments, Three.js (https://threejs.org/), and
apple-swift language with OpenGL-ES (tested in early 2017 at the beginning
of the project). In all tests performed Processing resulted in much higher
efficiency [29].
Part
of the idea of this project was to offer scientists a tool with which they
could analyze in three dimensions the data available in their office,
without needing a computer cluster or a high-efficiency computer. Thus the
tool was optimized to run on a medium-capacity personal computer. In the
final stages of the project, several tests were successfully carried out
to verify that the system could run on the scientists' computers.
In
the case of Plinius, the geography of Costa Rica was loaded through a CSV
file, which in turn was garnered from a GeoTiff type image, obtained from
the database of the PRIAS project of the National Center for High
Technology of Costa Rica (CeNAT http://www.cenat.ac.cr/en/prias-en/).
On
the other hand, the hypocenter positions were obtained from OVSICORI (http://www.ovsicori.una.ac.cr) and fed to the system through
a file of 140 000 rows, also of the CSV type. The data contained for each
row an event with its longitude, latitude, depth, magnitude, day, month,
year, etc.
The
only extra library used for the system was the
"processing.pdf" because the system is able to save files of
what was displayed in PDF format, all the rest of the features, including
interactive elements such as sliders and buttons were programmed
from scratch in Processing.
For
navigation in three-dimensional space, we used the system that our
research group iReal had already generated and tested in other projects,
which is a version adapted to the geography of the traditional navigation
scheme "virtual sphere" [26].
To
avoid occlusion problems the only rendering technique used
in visualization is wireframe, so geography, plate tectonic planes
or other objects do not hinder the observation of the
hypocenters behind them. This rendering technique also helps to
improve efficiency as it is a technique that requires low use of
computational capabilities.
Through the design and development process of
Plinius, we conducted several tests with a group of 
scientists from the Volcanological and Seismological
Observatory of Costa Rica (OVSICORI: http://www.OVSICORI.una.ac.cr).
Some user interactions (such as the three
points to define a seismic profile) are the conventional way to set these
configurations in the domain disciplines. Because of this, when users want to
identify one of these areas, Plinius takes an orthogonal position to allow the
usual data entry, since users are accustomed to working in two dimensions.
Other interactions, like the cylindrical filter, are entirely new in the domain
practices.
The obvious step forward for scientists using
Plinius is the possibility to see, for the very first time, a 3D visualization
of earthquake swarms with the options of zooming in and out and rotating in all
directions. Previously, they could only decide a profile or area of interest,
filter the data from the database, and visualize a section of this selected
data. This process took several hours, and in the end, they only saw a 2D cut —
if it was not what they were looking for, they would have to start the process
all over again.
We conducted a post-use qualitative evaluation
survey to compare the experience of Plinius to the traditional approach. Seismologists
were asked which approach they found more intuitive, what they would prefer to
use in the future, and which of the two methods they think is easier to use.
Most users and observers said that this tool is much more intuitive and, above
all, much faster than all the tools they use currently.
In all cases, we found that the visualization
was correctly used by the scientist, with minimal training required.
The aforementioned validated cases prove that
Plinius is a useful tool for scientists. Hypocenter data is naturally found in
three dimensions, yet traditionally, scientists have only been able to
visualize such data using 2D cuts.
Plinius now allows data to be seen as it truly
is, to scale, and how it naturally relates to other data such as geography,
positions of cities, volcanoes, and other important sites. Through Plinius, the
subduction of the Cocos Plate under the Caribbean Plate, as well as the Panama
Block can now be visualized in three dimensions (figure 12).
Fig. 12. Subduction of the Cocos Plate under
Caribbean Plate and Panama Block
Future prospects include adding ongoing
seismology research results to Plinius. One example is to incorporate lines to
reflect plate geometry, which would allow users to visualize the true geometry
of the tectonic plates.
While our tool responds satisfactorily to the
requirements defined by seismologists, an interesting suggestion is given by Ma
et al. [30]: adding to the tool a new dashboard that would allow users to have
statistical information such as scatter plots, box plots, and histograms. This
dashboard would be a complement to the visual part and would strengthen the
exploratory data analysis inside the system.
This approach is consistent with the work
developed in VisTravel [1], a system that includes eight views, several of
which depict statistical information. The primary objective of these systems is
to allow users a multi-perspective analysis.
The whole system was developed in a standardized way from the data
with its longitudes, latitudes, and depth, with the idea of making it easy
to adapt to other cases. In other words, the tool can visualize a set of
hypocenters anywhere in the world and continue to use filters, navigations
and other features immediately.
On the other hand, the whole project was developed in a
Public University in Costa Rica entirely with public funds, and therefore
it is open to collaborating in the adaptation of this tool or
its components with any other group of researchers that so wishes
or needs it.
We wish to thank the
Volcanological and Seismological Observatory of Costa Rica at National
University, OVSICORI (http://www.OVSICORI.una.ac.cr). We would also like to thank the scientists who provided
assistance as consultants, testers, and advisers.
Finally, we give thanks to the eScience
research program to which our iReal group belongs, which has supported this
project in all aspects of its administration and funding.
[1] Q. Li, Y. Wu, S. Wang,
M. Lin, X. Feng andH. Wang. VisTravel: visualizing tourism network opinion from
the user generated content. Journal of Visualization,
19(3), 489-502, 2016.
[2] F. Hernandez-Castro, J.
Monge-Fallas, M. Méndez-Morales, M. Protti-Quesada
Animation: Crustal Deformation in the Nicoya Peninsula
Associated with the September 5th, 2012 Earthquake. Scientific
Visualization. 2018;10(3): 29 - 47, DOI: 10.26583/sv.10.3.01.
[3] C. Kyriakopoulos, A. V. Newman, A. M. Thomas, M.
Moore‐Driskell
and G.T. Farmer, G. T. A new seismically constrained subduction interface model
for Central America. Journal of Geophysical Research: Solid
Earth, 120(8), 5535-5548, 2015.
[4] J. Meyer and T. Wischgoll. Earthquake
visualization using large-scale ground motion and structural response
simulations. In Scientific Visualization: The Visual
Extraction of Knowledge from Data (pp. 409-432). Springer, Berlin,
Heidelberg, 2006.
[5] W. Dzwinel, D. A. Yuen, K. Boryczko, Y. Ben-Zion,
S. Yoshioka and T. Ito. Cluster analysis, data-mining, multi-dimensional
visualization of earthquakes over space, time and feature space. Nonlinear Processes in Geophysics, 12,
117-128, 2005
[6] K. M. Fairchild, S. E. Poltrock and G. W. Furnas.
Semnet: Three-dimensional graphic representations of large knowledge bases. In Readings in information visualization (pp. 190-206).
Morgan Kaufmann Publishers Inc, January 1999.
[7] S. E. Palmer. Modern theories of Gestalt
perception. Mind & Language, 5(4), 289-323, 1990.
[8] X. Tao. Research on information visualization
design based on cognitive theory. In Design Management
Symposium (TIDMS), 2013 IEEE Tsinghua International (pp. 123-126).
IEEE, December 2013.
[9] A. Figueiras. Towards the understanding of
interaction in information visualization. In Information
Visualisation (iV), 2015 19th International Conference on (pp.
140-147). IEEE, July 2015.
[10] R. Fernandez and N. Fetais. Survey of Information
Visualization Techniques for Enhancing Visual Analysis. In Computer
and Applications (ICCA), 2017 International Conference on (pp.
360-363). IEEE, September 2017.
[11] N. Prapaitrakul and S. Phithakkitnukoon. EQviz: a
visualization tool for monitoring world earthquakes. In Adjunct Proceedings of
the 2015 ACM International Joint Conference on Pervasive and Ubiquitous
Computing and Proceedings of the 2015 ACM International Symposium on Wearable
Computers (pp. 1207-1211). ACM, September 2015.
[12] C. Willmesa, J. Weskamma, U. Baaser, K. G. Hinzen
and G. Bareth. SeismoGIS: A tool for the visualization of earthquake data.
In Proc. XXI ISPRS congress (Beijing, China, 3-11 July (Vol.
2008, p. 1239), 2008.
[13] R. H. Wolfe and C. N Liu. Interactive
visualization of 3D seismic data: A volumetric method. IEEE
Computer Graphics and Applications, 8(4),
24-30, 1988.
[14] D. Patel, C. Giertsen, J. Thurmond, J. Gjelberg
and E. Grøller. The seismic analyzer: Interpreting and illustrating 2d
seismic data. IEEE transactions on visualization and
computer graphics, 14(6), 1571-1578,
2008.
[15] G. Leonard, Z. Somer, Y. Bartal, B. Y. Horin, M.
Villagran and M. Joswig. GIS as a Tool for seismological Data Processing. In Monitoring the Comprehensive Nuclear-Test-Ban Treaty: Data
Processing and Infrasound (pp. 945-967). Birkhäuser, Basel,
2002.
[16] A. Chourasia, S. Cutchin and B. Aagaard.
Visualizing the ground motions of the 1906 San Francisco earthquake. Computers & Geosciences, 34(12),
1798-1805, 2008.
[17] R. H. Wolfe and C. N Liu. Interactive
visualization of 3D seismic data: A volumetric method. IEEE
Computer Graphics and Applications, 8(4),
24-30, 1988.
[18] B. T. Aagaard,T. M. Brocher, D. Dolenc, D.
Dreger, R. W. Graves, R, S. Harmsen and M. L. Zoback.. Ground-motion modeling
of the 1906 San Francisco earthquake, Part I: Validation using the 1989 Loma
Prieta earthquake. Bulletin of the Seismological Society of
America, 98(2), 989-1011, 2008.
[19] D. Patel, C. Giertsen, J. Thurmond, J. Gjelberg
and E. Grøller. The seismic analyzer: Interpreting and illustrating 2d
seismic data. IEEE transactions on visualization and
computer graphics, 14(6), 1571-1578.,
2008.
[20] L. Castanie, F. Bosquet and B. Levy. Advances in
seismic interpretation using new volume visualization techniques. first break, 23(10),
2005.
[21] M. Ivančić, Ž. Mihajlović and
I. Ivančić. Seismic data visualisation. In Information and
Communication Technology, Electronics and Microelectronics (MIPRO), 2015 38th
International Convention on (pp. 324-327). IEEE, May 2015.
[22] X. Liu, D. Li, Y. Xu and W. Xu. Research on 3D
Visualization Method of Seismic Data. International Journal
of Signal Processing, Image Processing and Pattern Recognition, 9(5), 441-454, 2016.
[23] D. A. Yuen, B. J. Kadlec, E. F.Bollig, W.Dzwinel,
Z. A. Garbow and C. R. da Silva. Clustering and visualization of earthquake
data in a grid environment. Visual Geosciences, 10(1), 1-12, 2005.
[24] J. W. Stoecker. TerraVis: A Stereoscopic Viewer
for Interactive Seismic Data Visualization, 2011.
[25] T. J. Hsieh. Understanding earthquakes with
advanced visualization. ACM SIGGRAPH Computer Graphics,
44(1), 4, 2010.
[26] F. Hernandez-Castro &
J. Monge-Fallas. Navigation Sphere: Optimizing Virtual Sphere for Terrains
Analyses. PONTE: International Scientific Researches Journal, 74(7), 2018.
[27] B. Saket, A. Srinivasan, E. D. Ragan and A.
Endert. Evaluating interactive graphical encodings for data visualization. IEEE transactions on visualization and computer graphics,
2017
[28] J. Monge-Fallas & F. Hernández-Castro.
An Intuitive 3D Interface for Defining Seismic Profiles by Plinius. PONTE: International Scientific Researches Journal, 74(4), 2018.
doi:
10.21506/j.ponte.2018.4.9
[29] F.
Hernandez-Castro, J. Monge-Fallas. Eficiencia
comparativa en animaciones en javascript. Tecnología en Marcha. 31(3), 142-149,
Julio-Setiembre 2018.
[30] X. Ma, D. Hummer, J. J. Golden,P. A. Fox R. M.
Hazen, S. M. Morrison and M. B. Meyer. Using Visual Exploratory Data Analysis
to Facilitate Collaboration and Hypothesis Generation in Cross-Disciplinary
Research. ISPRS International Journal of Geo-Information,
6(11), 368, 2017.
