International Journal of Geosciences, 2011, 2, 648-656
doi:10.4236/ijg.2011.24066 Published Online November 2011 (
Copyright © 2011 SciRes. IJG
Virtual Globes and Geological Modeling
Tsangaratos Paraskevas
Laboratory of En gi neeri n g G e ol o gy—Hydrogeol ogy, Sectio n of Geosci ences, Scho ol of
Mining and Metallurgical Engineering, National Technical University of Athens, Athens, Greece
Received May 27, 2011; July 9, 2011; accepted Septe mber 14, 2011
Virtual Globes such as Google Earth TM, revolutionize the way scientists conduct their research and the way
the general public use geospatial-related data and information. Improvement in the processing power and
storage capacities of computers, along with the increased Internet accessibility and connectivity, has sup-
ported the usage of Virtual Globes technologies. Even more, software releases of freely downloadable Vir-
tual Globes, such as Google Earth and NASA World Wind, has sparked an enormous public interest and in-
creased people’s awareness of spatial sciences. In this study, the Virtual Globes (VG) revolution is discussed
and a client-server Graphical User Interface (GUI) application is presented. The developed application en-
ables Google Earth TM Application Program Interface and activates spatial analysis, through enhanced
JavaScripts and Visual Basic script codes. The main scope was to present the methodology followed during
geological modeling along with the application capabilities when handling with data derived from digitized
geological maps and field measurements.
Keywords: Virtual Globe, Geological Modeling, Graphical User Interface, Spatial Analysis
1. Introduction
In every geosciences model that simulates a natural sys-
tem, interactive data manipulation and visualization, are
the most essential tools [1]. The projection of these pro-
cedures and their analytical results, are usually presented
by the form of digital or analogous thematic maps [2].
Digital or analogous, maps actually help users to under-
stand the spatial context of things, concepts, conditions,
process or events, simulating parts of the environment
and every day life. However, due to limitations of the
procedure following the construction of analogous maps
and in some extent 2D digital maps, this form of com-
munication among scientists and the general public is not
fully appreciated. Maps are flat, providing only a single
level of detail, they are static and the constructio n is usu-
ally slow [3]. Al Gore, former vise president of USA,
envisioned a “Digital Earth”, a multi-resolution, 3D rep-
resentation of the planet, into which we could embed
vast quantities of geo-referenced d ata [4]. Displaying the
real-world, either as a “mirror” of what we see, or as spe-
cialized display of social, economic, infrastructure or
environmental data. This form of representation over-
comes the limitations that an alogo us and 2D digital maps
own. By the end of the millennium, it was well estab-
lished the idea of a computer application, allowing users
to browser and search geographically indexed informa-
tion, projected on a cartographic representation of the
Earth in various scales and projections [5]. The materi-
alization of this idea was based on technological advan-
tages, such as orbiting satellites with spatial resolution of
a few meters, improvements in hardware and software
and the growth of broadband internet, where users have
World Wide Web (WWW) access without slow speed
and telephone line disruption [6]. Modeling a Virtual
Earth or a Virtual Globe (VG), for effective communica-
tion, enhancing science, distributing knowledge, under-
stand and insight phenomena and natural process, that
was the idea.
The advantages of implying such an idea are fully un-
derstood when trying to visualize geoscientific data. The
impact that those images have on viewers, are certainly
more direct and more powerful than those produced by
analogous and 2D static digital maps. Geology and ge-
ology-based scien ce are not an exceptio n. The geological
surface extent and lithological structure, the spatial dis-
tribution of faults and zones of discontinuities, the hy-
drographical network and other points of interest, are
basic knowledge, necessary for understanding and inter-
preting the geological model that represents the studied
Copyright © 2011 SciRes. IJG
area. The better they are visualized, the better they are
understood. However, the over mentioned processes of-
ten requires a working knowledge of image processing
and 3D modeling software and the ability to write script
codes from the side of the end-user.
The main scope of this study was to provide a meth-
odology through a well designed Graphical User Inter-
face application that enables the Virtual Globe technol-
ogy providing a helpful tool for handling geological data
in a more sophisticated manner.
2. VG Technology and Geological Data
VG are presented as 3D software models, which are ca-
pable of modelling the Earth or other environments of
the universe, like Moon, Venus and Jupiter. They pro-
vide the user with the ability to freely move around by
changing the viewing angle and position. Compared to a
convention al globe, VG have the add itional capability of
representing many different layers of information on the
surface of the Earth. These information layers concern
spatial and non-spatial attributes. Specifically, VG can
display, man-made features such as infrastructure net-
works and buildings, elevation data or representations of
demographic quantities such as population or outcrop
production. They pro vide the ability to browse maps and
geospatial data on the internet using the relative servers.
They can handle data such as geodatabases, shapefiles,
KML/KMZ, GPX and raster formats (JPEG, GeoTIFF,
MrSID) and integrate a wide variety of context like pho-
tos, videos, documents, 3D models and place them in a
geographic content. Recen t applications of 3D modelling
software, such as Google SketchUpTM have led to the
development of 3D earth science models that can be di-
rectly incorporated into VG scenes. However, unlike GIS
maps, VG display this information in photorealistic, life -
like three dimensions. Few of them can perform spatial
functions, e.g. visib ility, modelling , and prox imity search,
ArcGIS Explorer. Finally, add-ons and small extensions,
written in C#, VB or J# scripts can be loaded and com-
piled, adding new functionality to the programs.
One of the fields, in which VG technology have dy-
namically entered, is in the geoscientific domain. A gr eat
number of VG applications have been used in research
fields such as, global climate change, weather forecasting,
natural disasters (e.g. tsunamis, hurricanes), environment
conservation, travel, nature, people and culture, history
illustration, avian flu, online gaming, etc. [7]. As it can
be ascertained, map compilation and visualization of
geological related data, are undergoing a revolution in
the way they are performed. Assisted by Global Posi-
tioning System (GPS), Geographic Information System
(GIS), Web Map Services (WMS) and Virtual Globe
(VG) technology, the process and methods of geological
modelling have marked a major advance [8]. Consider-
ing the performance of various procedures that are im-
plemented for geological modeling through the use of
VG technology, [9] demonstrated a map deconstruction
process, named “GIS-ification”. During the process, the
analogue geological maps are transformed into interact-
tive maps and are introduced in Google Earth TM. Each
lithological unit of the map is scanned and introduced as
a separate layer having transparent mask. A similar step
by step process is described by [10], who presented a
relatively inexpensive and rapid method for creating a
3D model of geology from published quadrangle scale
maps and cross-sections, using Google Earth TM and
Google SketchUpTM software. Analogous procedure was
presented by [11]. Another method for coupling VG with
geophysical hydrodynamic models was presented by [1].
Firstly, users draw p olygons on the Goog le Earth screen.
These features are then saved in a KML file which is
read using a script file written in Lua programming lan-
guage. After the hydrodynamic simulation is being per-
formed, another script file is used to convert the resulting
output text file int o a KML file for visualization, where t he
depths of inundation are represented by the colour of dis-
crete point icons. The visualization of a wind speed vector
field was also included as a supplementary example.
All of those procedures share the same characteristic.
They are developed by Earth scientists in order to effect-
tively distribute their research outcomes to both, other
scientists and the general public. However, at the present
time, the available software has limitations when it
comes to mapping and modeling features such as geo-
logical orientation data, dip and dip directions values for
faults, and geological structure features. In ad dition, they
do not perform spatial nor sta tistical functions over areas
of interest, limiting the usage of the VG, to be a tool for
data input and visuali zat i on.
The main scope of this study is to develop a form of
client-server GUI application, transforming the static
GIS-based applications into a more dynamic tool for data
input, manipulation and visualization. A GUI that pro-
vides the end user, with spatial analysis functions, having
also the ability of presenting data in a familiar spatial
context, easily to be found online and accessed any time
and in any place, though VG technology.
3. Methods and Material
It is well known that there has been an increasing num-
ber of amazing technologies that can boost our software
application to a higher level of power and usability. Ajax
(Asynchronous JavaScript +XML) model is one of them
[12,13]. It seems that these technologies have renewed
Copyright © 2011 SciRes. IJG
the main objective of the Web development community
which is to create applications that are powerful and re-
sponsive, like any desktop based application, having
though the advantage of accessibility via the WWW.
With Ajax, web applications can retrieve data from the
server asynchronously in the background without inter-
fering with the display and behaviour of the existing
page. The use of Ajax techniques has led to an increase
in interactive and dynamic interfaces on web pages.
Google Earth API, Google Maps and its Java Script API,
are free services, allowing anyone to build sophisticated
3D map applications that embraces the Ajax model. The
Google Earth API can display placemarks, lines, poly-
gons, overlays and 3D models, on the imagery just as the
standalone version of Google Earth can. The plug-in
supports several of the Google Earth layers of informa-
tion, including terrain data, infrastructure network, bor-
ders and buildings.
The Google Earth API can be used to develop dy-
namic Client Server and Web based applications. In this
study, a Client-Server Application (CSA) was devel-
oped. A typical CSA describes an application architect-
ture in which the client requests an action or service from
the provider of service, the server. Considering a Web
browser and a Web server, when we address a URL in the
browser window, it requests a page from a Web server.
The server returns an HMTL page to the client, which
parses the page (data) and displays it on the computer. In
this study, the role of the client is played by the devel-
oped GUI, while the Google Web Server acts as the re-
quired Web server that returns to the client requested
data. The developed GUI is a CSA for displaying very
large, global scale, terrain models. The terrain database is
stored on a Google Web Server and the client program
fetches the data required for generating an image on our
screen with the wanted resolution. As moving around,
the program loads and displays data as needed.
3.1. Data Format
The process of data input using Google Earth, adopts
Keyhole Markup Language (KML) as many of the avail-
able VG do. KML is used as the format for describing,
organizing and visualizing geographical objects [14]. It
represents a hierarchical data system where the geo-
graphical objects can be populated in a n ested tree struc-
ture [15]. There are mainly three types of methods that
can be used to handle geospatial data with the use of
KML format [7]:
using binary values, that can be stored and organized
as KML elements namely: points, linestring, linear-
ring, polugon, multigeometry,
using image in the format of png, jpeg, gif, that can
be introduced as KML elements namely: photoOver-
lay, screenOverlay and groundOverlay,
using 3D models, which can be modeled base on bi-
nary values, images, Sketchup models, etc.
The description for every object is provided by fol-
lowing a certain structure, using pairs of KML tags, in
which, information about the number of objects, their
coordinates and style, is presented. For displaying a point,
the KML file takes the following form:
<Point>, opening tag of Object
<name> TestPoint_A </name>, naming the point
<coordinates> 24.015, 38.174 </coordinates>, speci-
fying the coordinates of the object
</Point>, closing tag of Object
KML’s full abilities are presented when using 3D mod-
els, combined with COLLADA (COLLAborative Design
Activity) model. COLLADA is defined as an open stan-
dard XML schema for exchanging digital assets among
various graphics software applications, usually id entified
with a “.dae” (digital asset exchange) filename extension.
Models are constructed and defined independently of the
Google Earth application, having their own coordinate
space and using applications such as SketchUpTM, 3D
Studio Max, Softimage XSI or Maya. However, when a
3D model is imported into Google Earth, it is translated,
rotated and scaled to fit in to the Google Earth coordinate
As already mentioned, many software that enables VG
technology, have the ability to handle ESRI shapefiles
and import simple text files which refer to geospatial
data. To upload a point ESRI file, the first action is to
transform the file into a plain text format. The next thing
is to parse the needed information, by applying a simple
rule searching script and retrieve the corresponding value,
which has to do with coordinates, altitude and name or
ID of the observation point. Finally, the text’s structure is
reformed and saved in a KML format.
3.2. System Architecture and Main
The developed CSA was designed and compiled using
Visual Basic programming language, enabling the Google
Earth Application Program Interface. It consists of a sin-
gle form that contains all the specific scripts and codes
that are capable of performing spatial and statistical analy-
si s functions .
The form contains a split panel control oriented verti-
cally. The left part contains the controls that are used to
add and manipulate the available data. There are two sets
of controls in this panel. The first set contains all the
necessary controls for the data input and the data trans-
formation procedures. The second set of controls is de-
Copyright © 2011 SciRes. IJG
signed to perform spatial and statistical analysis, from
the available or the requ ested data. The right panel of the
control contains a WebBrowser Object, docked into it.
This control contains objects such as HTML Document
or others. In the case presented, the HTML Document was
the Google Earth API. Events that occur in the Google
Earth API which is contained in the WebBrowser Object
are captured and forward to a procedure in the left panel
of the Form. A Visual Basic script which is embodied in
the form’s source code, executes specific functions in
response to the forwarded Events. The same procedure is
executed with Events and Scripts performed in the Form
and forwarded to Google Earth HTML Object (Figure
Google Earth API does not include the option of up-
loading a KML file from the local disk of a user, mainly
for security reasons set by the Internet Explorer protocol.
There are few ways to bypass this obstacle, the latest of
which is adopted by the developed GUI:
Writing a proxy that uses an HTML Form to upload
the file to a server and then make it available on the
local machine.
Changing the security settings of the HTML object by
adding it to the “trusted sites” and changing it from
“disab le” to “prompt”, th e “initialize and script Ac tiveX
controls not marked safe for scripting” checkbox. This
option uses the ActiveXObject (“Scripting. FileSyste-
Getting a string of the local KML file and pass this
string to the parse KML method of the API.
One of the main features of the developed VG is its
ability to handle orientated geological data and present
them in a way they don’t loose their spatial characteris-
tics. The KML file that references the COLLADA file
contains a Lat-Long insertion point and information con-
cerning orientation, and scale information. The KML file
also contained information about the lithological unit in
which each observation was placed. With the use of a
few script lines introduced in the source code of the GUI,
the dip/dip direction icon was dynamically orientated
relative to the geographic north, so that when the view
orientation was changed, the dip/dip direction icon ro-
tates, so as to maintain a constant azimuth. A similar
code can be found in the paper presented by [16], how-
ever in the developed GUI, the user is able to introduce
an orientated icon without producing any programming
code. The user simply imports the icon referring to the
field measurements that are introduced into the form of
the GUI.
The developed VG embodies additional functions and
spatial analysis tools, such as 3D distance measurement,
slope calculation, terr ain section an alysis, know ledg e and
information that help the users understand the actual geo-
graphical and geological environment. Specifically, the
user has the ability to profile the surface and export screen-
shots, as an image file but also to export the elevation
data into a text format file, for further spatial calcula-
3.3. Research Area
The study area concerns the wider area of the Perfection
of Xanthi, ap proximately 800 Km2, bounded to the north
by the Greek-Bulgarian borders and extended to the
south up to the Neogene Thrace basin (Figure 2). The
geological structure consisted of a marble unit (marbles
and schists), a gneissic unit (magmatites, gneisses, am-
phibolites and ultra maf ic ro c ks) o f Palaeo zoic ag e an d th e
Tertiary mollasic and igneous rocks. The elevation values
of the area, varied between 30 to 1800 m approximately.
However, from a morphological point of view, the area is
characterized as mountainous. The available data came
by digitizing the prev ious conducted geo logical map [17],
at 1:50,000 scale as well as by field observations that
represent the geological structure.
Figure 1. The general schema design.
Figure 2. Research area.
Copyright © 2011 SciRes. IJG
Each GIS geological map is represented by a number
of ESRI shapefiles (Figures 3 and 4). Specifically, the
files contain: geological arcs and polygons shape files
(.SHP), geological symbols shape files (.SHP), base map
image file (.tif, .jpeg, .bmp). Each of the projected layers
was converted from EGSA’87, Greek Geodetic Reference
System (1987), to WGS84 Geographic Coordinate Sys-
tem. During the implementation, named and scaled place-
mark icons were placed in order and exported as KML
3.4. Evaluating the Developed Application
In order to test the efficiency of the developed GUI, the
case of visualizing the field observations was investi-
gated (dip and dip direction) while evaluating the pro-
duced images and projections for the tectonic feature that
was plotted.
Dip and dip direction, refer to the orientation of a geo-
logical observation point. The dip direction of a fault, or a
joint, is projected as a line, representin g the intersec tion o f
the observation point with a horizontal plane and it is
characterized by a positive angle between 0˚ - 360˚. On a
static 2D geological map, this is represented by a short
straight line segment oriented, parallel to the dip direc-
tion line. The baseline orientation is north but this could
be true north, grid north or magnetic north.
A dip value is actually the steepest angle of descent of
a titled observation point relative to a horizontal plane
and is characterized by a positive angle between 0˚ - 90˚.
The symbol used in a static 2D map, is a short line at-
tached to the symbol of the dip direction, pointing in the
direction which the planar surface is dipping down. As
mentioned by De Paor et al. 2010, Google Earth by de-
fault handles placemark icons, as dip and dip direction
are, dynamically orientated relative to the viewer, creat-
ing though a false visualization of the observation point
when the user choose to change the view angle.
In the developed GUI, the dip and dip direction values
are imported through the left panel while after been
transformed into KML file they are forward to the right
panel, the WebBrowser Object. The HTML Document
which contains the Google Earth API, also includes action
buttons that, when clicked, perform certain functions. In
the case of placing the oriented placemark icons the
function is written using Javascript code as the one pre-
sented in Figure 5.
The outcome of this procedure is the projection of the
oriented observatio n poin ts in su ch a manner that it cou ld
produce a true visualization of the observation point,
even when the user changes the view angle (Figure 6).
A similar procedure was adopted for placing geological
Figure 3. Geological layer overlapping 3D terrain data.
Copyright © 2011 SciRes. IJG
Figure 4. Fault layer overlapping 3D terrain data.
Figure 5. JavaScript code for orientated 3D model.
Figure 6. Visualization of the orientated icon from different
Copyright © 2011 SciRes. IJG
cross-sections, overlapping the 3D terrain in their correct
position and orientation. The geological crosssection,
thought as an interpretation of subsurface geology, are
designed using Google SketchUpTM, exported as COL-
LADA script file and positioned in Google Earth API
using specific KML files.
The GUI enabled also certain spatial analysis functions to
measure distance, extract altitude information, perform
slope calculation and produce elevation profile (Fi gure 7).
Figure 8 shows the script code for extracting the alti-
tude information of a given area defined by x and y co-
ordinates. In this case the needed altitu de information are
obtained from the Google Web Server, passed to the HTML
Document and forward from the right panel, WebBrowser
Object to the left panel and the B set of controls for fur-
ther spatial and statistical analysis. Through enhanced
visual basic codes the raw elevation data were analyzed
for each geological layer introduced by the KML file.
The GUI parse the KML file and calculates for each
layer the maximum and minimum elevation point, the
mean elevation value and the surface extent for each
4. Discussion and Conclusions
The ability to manipulate geospatial o bjects makes VG a
valuable tool for many research fields and scientific ap-
plications. They are presented as a powerful tool and an
advanced platform for scientific visualization among
decision makers, researchers and the general public. Their
specific characteristics, provides significant opportunity
for the science community to communicate information
and share the results o f often complex models within the
community, and general public that may could not oper-
ate and access spatial technologies such as GIS, remote
sensing and visualization products [18]. According to
[19], VG embody features and functionality that provide
significant advantage over traditional spatial mapping
interfaces, since:
the earth imagery displayed on a globe structure is
free of distortion,
data displayed on VG can be viewed at any scale and
from any angle,
VG provide a large degree of interactivity, allowing
the user to move to different locations and visualise
different type of spatial data.
Geologists and geological organizations worldwide are
investigating web technologies, to upload relative data
that have been converted from analogue to digital spatial
format, vectorized linework and attributed polygons. VG
provide an informational environment in which one can
integrate multiple data layers onto one interactive interface.
This process provides the capability of visualizing critical
information at an instance.
Figure 7. Elevation profile.
Copyright © 2011 SciRes. IJG
Figure 8. JavaScript code for extracting altitude values.
The advantages of the developed client-server applica-
tion are summarized as follows:
easy to handle by non-GIS users,
open source licence,
provides geographic orientation and panning/zoom-
ing capabilities ,
statistical analysis,
spatial feature extraction,
visualizing critical information as an instance.
However, although VG could be characterized as an
outstanding innovation in the field of visualizing informa-
tion, it still possesses some disadvantages:
not applicable if intern et connection fails or with low
in the present time it depends on other software to
convert into app ropriate format the needed geospatial
data, i.e. digitizing the geological maps.
It’s clear that greater magnitude of advantages out-
weighs its disadvantages. Th e initial envision of a “Digi-
tal Earth” could be finally projected as many well de-
signed Graphical User Interface applications that en-
hance spatial analysis and most of all, interact with VG
technology and WWW services, are developed. These
sophisticated GUI transforms the static GIS-based ap-
plications into a more dynamic tool for data input, ma-
nipulation an d vi sual i zat i on.
Future improvement in the compatibility between GIS
and VG technologies will change the way in which spa-
tial data can be used to inform natural resource manage-
ment and land use planning.
5. References
[1] N. Q. Chien and S. K. Tan, “Google Earth as a Tool in
2-D Hydrodynamic Modeling,” Computers & Geo-
sciences, Vol. 37, No. 1, 2011, pp. 38-46.
[2] D. Tiede and S. Lang, “Analytical 3D Views and Virtual
Globes—Scientific Results in a Familiar Spatial Con-
text,” ISPRS Journal of Photogrammetry and Remote
Sensing, Vol. 65, No. 3, 2010, pp. 300-307.
[3] M. F. Goodchild, “Cartographic Futures on a Digital
Earth,” Cartographic Perspectives, Vol. 36, No. 4, 2000,
pp. 3-11. doi:10.3138/5717-2P33-0332-7364
[4] A. Gore, “The Digital Earth: Understanding our Planet in
the 21st Century,” Photogrammetric Engineering and Re-
mo t e S ensing, Vol. 65, No. 5, 1999, pp. 528-530.
[5] J. Bailey, “Virtual Globes,” In: B. Warf, Ed., Encyclope-
dia of Geography, Sage Publications Inc., Los Angeles,
2010, pp. 3528.
[6] J. Bailey and A. Chen, “The Role of Virtual Globes in
Geoscience,” Computers & Geosciences, Vol. 37, No. 1,
2011, pp. 1-2. doi:10.1016/j.cageo.2010.06.001
[7] A. Chen, G. G. Leptoukh, S. J. Kempler and L. Di,
“Visualization of Earth Science Data Using Google Earth
TM,” The International Archives of the Photogrammetry,
Remote Sensing and Spatial Information Sciences, Vol.
34, Part XXX, 2009.
[8] S. J Whitmeyer, J. Nicolletti and D.G. De Paor, “The
Digital Revolution in Geological Mapping,” GSA Today,
Vol. 20, 2010, pp. 4-10. doi:10.1130/GSATG70A.1
[9] D. G. De Paor, C. Simpson and S. Whitmeyer, “Decon-
structing Classical Geologic Maps Using Google Earth’s
Keyhole Markup Language,” Geological Society of
America Abstracts with Programs, Vol. 40, No. 6, 2008,
p. 348.
[10] G. J Walsh, “A Method for Creating A three dimensional
Model from Published Geologic Maps and Cross Sec-
tion” U.S. Geological Survey, Open-File Report, 2009, p.
16. http://pubs.usgs. gov /of/2009/1229.
[11] C. Simpson and D. G. De Paor, “Restoring Maps and
Memoirs to Four-Dimensional Space Using Virtual
Globe Technology: A Case Study from the Scottish
Highlands,” In: R. D. Law, R. W. H. Butler, R. E. Holds-
worth, M. Krabbendam and R. A. Strachan, Eds., Conti-
nental Tectonics and Mountain Building: The Legacy of
Peach and Horne. Geological Society of London Special
Publication, Vol. 335, pp. 427-439.
[12] A. Lewis, M. Purvis, J. Sambells and C. Turner, “Begin-
ning Google Maps Applications with Raiks and Ajax:
from novice to professional,” Springer, Berlin, 2007, p.
384. doi:10.1007/978-1-4302-0380-3
[13] Z. Li, C. Yang, H. Wu, W. Li and L. Miao, “An Opti-
mized Framework for Seamlessly Integrating OGC Web
Services to Support Geospatial Sciences,” International
Journal of Geographical Information Science, Vol. 25,
No. 4, 2011, pp. 595-613.
[14] T. Wilson, D. Burggraf, R. Lake, S. Patch, R. Martell, B.
McClendon, M. Jones, M. Ashbridge, B. Hagemark, J.
Copyright © 2011 SciRes. IJG
Wernecke and C. Reed, “OGCs KML. Open Geospatial
Consortium, Document OGC07-147r2”, 2008, p. 251.
[15] J. Wernecke, “The KML Handbook: Geographic Visu-
alization for the Web,” Addison-Wesley, Upper Saddle
River, 2009, p. 368.
[16] D. G. De Paor and S. J. Whitmeyer, “Geological and
Geophysical Modeling on Virtual Globes Using KML,
COLLADA, and Javascript,” Computers and Geo-sci-
ences, Vol. 37, No. 1, 2011, pp. 100-110.
[17] P. Ilias, D. Rozos, G. Konstandopoulou, E. Dimadis, S.
Salappa, E. Apostolidis and A. Gemitzi, “Engineering
Geology Study of Disastrous Phenomena in Central-
Rhodope Mountains,” Institute of Geology and Mineral
Exploration, Mandelbachtal, 2000, p. 174.
[18] J. P. Aurambout, C. Petti and H. Lewis, “Virtual Globes:
the Next GIS?” Lecture Notes in Geoinformation and
Cartography, Landscape Analysis and Visualization, In:
W. Cartwright, G. Gartner, L. Meng and M. P. Peterson,
Eds., Springer, Berlin, 2008, pp. 509-532.
[19] A. Riedl, “Digital Globes,” In: W. Cartwright, M. P. Pe-
terson, G. Gartner, Eds., Multimedia Cartography, Sprin-
ger, Heidelberg, 2007, pp. 255-266.