GIS Application for Solar Potential Estimation on Buildings Roofs

Matej Brumen, Niko Luka c, Borut

Zalik
University of Maribor,Faculty of Electrical Engineering and Computer Science,
Smetanova ulica 17, Maribor, Slovenia
emails: {[email protected], [email protected], [email protected]}
AbstractOver recent years, photovoltaic systems have become
increasingly popular for the installation on buildings rooftops.
One of the main questions for investors is to nd which rooftop
is suitable for photovoltaic system installation in order produce
maximum electrical energy. This paper presents Geographic
Information System (GIS) for the solar and photovoltaic
potential estimation for a whole municipality. The presented GIS
consists of a server environment, web application, and mobile
application. The developed application was successfully deployed
within a regional project for the Municipality of Beltinci in
Slovenia.
Keywords-GIS; Solar Potential Estimation; Web Application;
Geoserver; LiDAR
I. INTRODUCTION
Solar energy is a renewable energy source that can be
converted into electrical energy using the photovoltaic (PV)
effect. Since the advancement of PV technology buildings
rooftops have become suitable regions for generating electric-
ity from the received solar irradiance (i.e., radiation incident
on a surface). Of course, not all rooftops are appropriate for the
installation of PV systems due to unsuitable topography and
shadowing from the surrounding environment. Determination
of a rooftops suitability can be done using a manual survey
by experts in this eld. However, this process is highly unsuit-
able for the estimation of Solar Potential (SP) over a larger
region with thousands of rooftops. An automated approach
is necessary for larger regions. Such approaches produce
results that are then displayed using web-based Geographics
Information System (GIS) applications which allow users to
overview SP estimation for their buildings rooftops. There
are many web applications that allow estimation of SP and
each application displays results that are produced using a
specic solar irradiance estimation method. Moreover, some
applications allow the user to view the PV installation costs
and the revenue over a few years regarding a given rooftop.
This paper presents a novel GIS for the SP estimation of
rooftops. The system consists of a server and client applica-
tions. All data, including the calculated solar radiation data is
stored on the server side of the environment. SP estimation
is based on a newer method that considers analysing the
meteorological data of solar irradiation and the simulation of
shadowing over the classied Light Detection And Ranging
(LiDAR) data using the method presented in [1] [2]. Further-
more, the shadowing from nearby vegetation and surrounding
terrain is considered for SP evaluation. LiDAR is a remote
sensing technology that emits laser impulses for determining
the positions of objects on the landscape [3]. An unstructured
3D point-cloud of the landscape is obtained as a result of the
scan. Therefore, the local topography can be estimated. The
presented GIS is based on a newer SP estimation method, and
provides more accurate insights into the suitable surfaces for
PV systems installation.
In the 2nd section, the methods for SP estimation and some
web applications for viewing/calculating SP are presented. In
the 3rd section, our method for SP estimation is presented to-
gether with the technologies used for server-side environment,
web application and mobile application development. In the
4th section, the results of our methods and applications are
shown and are concluded in the 5th section.
II. RELATED WORK
A. Solar Potential Estimation
The use of geospatial data for automatic solar potential
estimation is the rst step towards accurate solar potential
estimations for buildings rooftops within urban areas. Sev-
eral different models have been suggested [4][9] that use
topographical characteristics of the Digital Elevation Model
(DEM). Ruiz-Arias et al. [10] conducted a detailed comparison
of the better-known models for the solar potential estimation,
and reported the r.sun model [9] as being one of the most
accurate. Voegtle et al. [11] analysed the solar potential on
rooftops in urban areas. Kassner et al. [12] represented LiDAR
as 2.5D point-cloud in order to estimate solar potential with the
help of masked roof contours. Jochem et al. [13] introduced
the method for automatic roofs planes detection in the 3D
point cloud by using an implementation for solar potential
estimation based on [9]. Their method uses the clear-sky index
(CSI). Hoerka and Kanuk [14] also used the r.sun model for
solar potential estimation within urban areas. Jochem et al.
[15] introduced transparent shadowing caused by vegetation
when calculating the solar potential. They considered the ratio
between the rst and the last echos of the laser pulses for the
measurement of transparency. Such vegetation shadowing can
only be used during the specic season when the scanning was
performed. A more accurate approach was used by Levison et
al. [16] with the use of high resolution ortophotos and LiDAR,
together with the vegetation growth model, in order to predict
the impact of rooftop shading and solar irradiance. The authors
had manually surveyed and classied the data for accurate
1 Copyright (c) IARIA, 2014. ISBN: 978-1-61208-333-9
WEB 2014 : The Second International Conference on Building and Exploring Web Based Environments
simulation of the vegetation growth. Hence, only smaller urban
areas could be analysed in reasonable time.
B. GIS Applications For Solar Potential
There are many web-based environments that allow the
user to overview the irradiance or PV estimation of the rooftops
over a local region. Comparisons between different features
of these applications are shown in Table [I]. The considered
features are:
Cell colouring: if the application provides an additional
layer, where the cells are coloured with the intensity of solar
radiation by using a colouring scale.
Monthly PV details: if the application considers further
information about the monthly solar potential values for the
selected building.
Investment plan: if the application provides any investment
plans for the users that want to install solar panels on their
rooftops.
Rooftop details and ROI: if the application provide addi-
tional information about the selected rooftop or its selected
part via user-selectable Region Of Interest (ROI). Some
application partially support this by providing only one of
these two features.
User parameters: if an application supports entry elds for
user specied parameters (rooftop angle and orientation,
rooftop usability percentage).
Rooftop angle consideration: if the application considers
rooftop angles during the solar potential estimation. This
could be obtained from custom user parameters or extracted
from LiDAR point cloud.
Some applications provide further analysis, such as econom-
ical investment planning over the selected rooftops, whilst
other types of applications display coloured cells as a new
raster/vector layer over the rooftops or the entire region. Few
display the measured data of the existing PV systems installed
on rooftops.
Figure 1. Overview of the solar potential vector layer over the rooftops using
the Mapdwell solar tool [27].
There are different types of data representation in GIS-
based applications. The two more advanced applications from
Table I were considered for a more detailed comparison.
These applications were Mapdwell Solar Tool [27] [28] and
NYC Solar Map [23]. Both applications support solar potential
estimation for selected parts of the rooftop and the calculation
for a detailed investment plan. The main difference is that
Mapdwell Solar Tool has a visual representation using a vector
layer (i.e., pixels coloured by intensity) of the solar potential
values (see Figure 1). This feature has an visual advantage
for quickly nding which parts of the rooftops are the most
suitable. However, this is more difcult to achieve on NYC
Solar Map because the user has to rst select each rooftop to
acquire detailed information of the solar potential estimation,
whilst not having a more detailed visual overview as to which
parts are more suitable (see Figure 2).
Figure 2. Detailed overview of the solar potential for the selected rooftop
area on the NYC Solar Map [23].
The presented GIS application supports both the solar
potential raster layer and the detailed per-building description,
and is explained in more detail within the next sections.
III. SOLAR POTENTIAL ESTIMATION AND GIS
ENVIRONMENT
The rst subsection briey summarises the theoretical
background of the used solar potential method, whilst the fol-
lowing three subsections present the developed GIS application
that consists of server, web, and mobile environments. The
entire workow of the presented GIS, reecting the considered
design choices, can be seen in Figure 3.
Figure 3. Workow of the proposed GIS, where normal rectangles represent
services and applications, while rounded rectangles represent the data.
A. Solar Potential Estimation
Solar potential is estimated using a newer method [1],
where multiresolutional shadowing is considered as well long-
term irradiance measurements by a pyranometer. Heuristic
variable shadowing from vegetation is also considered in the
estimation.
At rst the LiDAR data is arranged into a regular grid
consisting of 2.5D cells. The size of each cell is set at 1m
2
. The
height of the cell is dened by the highest point of the points
2 Copyright (c) IARIA, 2014. ISBN: 978-1-61208-333-9
WEB 2014 : The Second International Conference on Building and Exploring Web Based Environments
TABLE I. COMPARISON OF DIFFERENT GIS APPLICATIONS FOR THE SOLAR POTENTIAL ESTIMATION.
Application/Feature Cell colouring Monthly PV details Investment plan Rooftop details and ROI User parameters Rooftop angle consideration
Anaheim [17] YES NO NO NO NO NO
Berkley [18] NO N/A N/A NO NO N/A
Boston [19] YES YES YES YES YES NO
Denver [20] NO YES NO NO NO NO
Los Angeles [21] YES NO NO Partial NO NO
Madison [22] YES YES YES Partial YES YES
NYC [23] NO YES YES YES NO YES
Salt Lake City [24] YES YES NO Partial YES YES
San Diego [25] YES NO NO Partial NO YES
San Francisco [26] NO NO NO Partial NO NO
Mapdwell (Cambridge) [27] [28] YES NO YES YES NO YES
Dach acheneignung f ur Solaranlagen [29] Partial NO NO Partial NO N/A
Bristol [30] Partial NO NO Partial NO N/A
Wien [31] YES NO NO Partial NO N/A
within a cell. The daily solar irradiance is then calculated as:
I(t) =

sw
se
(I
b
(t)(1 S(t)) + I
d
(t))[Wm
2
], (1)
where t is the calculation time-step, se and sw are sunrise
and sunset, I
b
the calculated direct radiation, I
d
the calculated
diffuse radiation and S(t) [0, 1] the shadowing factor. Time
and space-dependent shadowing depends on an accurate posi-
tioning of the Sun, which can be obtained with the Almanac
algorithm [32]. The per-cell irradiance is calculated by con-
sidering many topographical relationships, such as slope and
orientation. The solar potential is then computed as cumulative
yearly irradiance in Watt hours per square meter [Whm
2
]:
P(t) =
365

t=1
I(t)[Whm
2
]. (2)
Additionally, P(t) can be multiplied by the constant ef-
ciency factor for a given semi-conducting material in order
to estimate approximate electrical energy production. The
constant efciency is usually taken as the peak-efciency of
a given semiconducting material depending on the type of
PV module. Recently, newer methods [33] [34] are being
developed for more accurate PV potential estimation.
Once the calculation is complete, the output data is con-
verted into the Environmental Systems Research Institute
(ESRI) shape le, where each cell location is stored together
with the values of the estimated solar potential.
B. Server Side Of The Environment
Geographical information systems are available as stand-
alone desktop applications (Quantum GIS [35], The Integrated
Land and Water Information System (ILWIS) [36]) or network
services (GeoServer [37], Mapnik [38]) running on remote
computers. Server applications support Web Map Service
(WMS) and Web Feature Service (WFS) requests, where
results are returned as WFS or WMS layers to the client-side
applications that process the requested data.
The GeoServer and PostgreSQL applications are used
on the server-side. GeoServer [37] is an open-source server
solution written in Java. The GeoServer allows the manage-
ment of the WFS and WMS requests that are specied by
Open Geospatial Consortium (OGC) [39]. Geospatial data is
stored in the PostgreSQL [40] database with a PostGIS [41]
extension. Data from SP estimation are converted into an ESRI
Shape format (SHP) and imported into the database that can
be accessed from the GeoServer. A proper colouring technique
is required because no colour value is stored in the solar shape
le. The GeoServer provides the OGC standardised Styled
Layer Descriptor language (SLD) that is used for customising
the shape colors according to the specied attribute values.
Colouring of the shapes is divided into 6 groups.
Each group has its own colour, e.g., groups group
0
to
group
5
are coloured in black, blue, cyan, green, yellow, and
red. The initial value of group
0
is min(solarpotential) and
the value of group
5
is max(solarpotential). The rest of the
key values are calculated as:
v
i
= (P
max
P
min
) 0.2 + v
i1
. (3)
The remaining values are interpolated.
C. Web Application
A web-based client application is necessary in order to
visualise the stored data and the pre-calculated solar potential.
Therefore, a good solution is to create a web-based client
application that consists of a map view for showing ortophotos
and the solar potential raster layer of the rooftops within a local
region. The simple point-in-polygon test can be performed by
using the combined data of the estate vector maps together
with their address information. The solar map is then aligned
to the vector map, as shown in Figure 4.
Figure 4. Alignment of the cells (left: building shape; middle: disarranged
cells; right: arranged cells).
When using this method, all the solar cells can be linked
to the rooftops and simple requests for each rooftop can be
made. Hence, the users can select the rooftops to see detailed
3 Copyright (c) IARIA, 2014. ISBN: 978-1-61208-333-9
WEB 2014 : The Second International Conference on Building and Exploring Web Based Environments
graphs of the cumulative calculated solar irradiance throughout
the year. OpenLayers [42] is used for displaying the necessary
layers via WMS. The WFS requests are used for obtaining
geometry and attribute data. The result of these requests is the
output formatted with Geometry Markup Language (GML),
which can be parsed with OpenLayers API. The JQuery [43]
libraries are used for Graphical User Interface (GUI) with
which simple pop-up windows can be created for displaying
details. Moreover, a Portable Document Format (PDF) report
can be obtained for the detailed information of a given rooftop.
PDF reports were generated by using L
A
T
E
X interpreter.
D. Mobile Application
Mobile devices are increasing in performance and support-
ing many hardware and software features. Therefore, a mobile
application can provide the same amount of information as
the web-based application, whilst it also supports the nearest
object search with the use of Global Positioning System (GPS)
technology. The developed application is supported on devices
with Android 2.2 or a higher version. The tabbed view is used
to categorise information about the calculated solar potential
because many of the devices still have low-resolution displays.
The map is shown on the rst tab, where the users can
navigate through the area and look for suitable roofs. With
the simple selection of the rooftop, the detailed data about
that rooftop can be viewed within the second tab. On the
map tab the WebView component is used where the mobile
version of the presented web-application is loaded, which is
a simplied version targetted for mobile devices. Using the
Javascript interfaces that are enabled on Android, the Java
and Javascript code can be connected together. For example,
methods implemented in Java (Android Application) can be
accessed from Javascript le on the web. This technique is
used for the data transfer when the user taps on the rooftop.
Application supports the GPS search of the nearest roof, and
also considers the devices orientations. Then it transmits the
GPS coordinates and orientation values to the server in order to
nd the closest rooftop. When the search is completed on the
server, the ID of the building is transferred back to the mobile
application, and the building with received ID is selected.
IV. RESULTS
Solar potential estimation and GIS-based application were
successfully applied for the Municipality of Beltinci in Slove-
nia, where the estimation was performed over 5605 rooftops.
The deployed web application is accessible at [44]. The mobile
application can be obtained at Google Play under the name
SolarEnergo, which currently only supports the Slovene lan-
guage. The total solar potential for all the considered rooftops
was estimated at 181 GWh per year, and the PV potential was
within a range from 16 GWh to 26 GWh, depending on the
type of PV material. The rst step was to obtain the classied
LiDAR data (see Figure 5), and to estimate the solar potential
for each rooftop. The considered solar potential estimation
method also required the use of long-term diffuse and direct
irradiance data from the nearby meteorological station.
Figure 5. Visualisation of the regular grid created from point cloud of LiDAR,
which is classied data into buildings, vegetation, and terrain.
The solar potential was estimated over a regular grid that
was created from LiDAR point-cloud. The grid consists of
multiple cells where each cell contains position, solar potential,
and approximate PV potential based on constant efciency
characteristics for amorphous and polycrystalline cells. In the
second step each cells information was prepared and converted
into a GeoServer supported format. The output was converted
into multiple shape les (see Figure 6) from which the PDF
reports were generated. All of the data was then inserted
into the database that can be accessed from the developed
applications via GeoServer. The colour of each cell was based
on its solar potential value, where the style was dened by
GeoServer.
Figure 6. Graphical overview of the solar potential values for the same region
of the WMS layer on GeoServer (left) and the regular grid of the LiDAR data
(right).
When a user selects a given rooftop, the detailed data
is shown in the pop-up window. These data consists of a
buildings address, the total solar potential throughout the
year, and the approximately produced electricity. There is an
additional graph where the data for each month are displayed
(see Figure 7). SP and PV values in the graph are represented
by different colors. The brown colour represents PV potential
for A-Si material, red is for P-Si material and blue colour
represents the solar potential.
HTML5 supported browser is necessary for this functional-
ity. Detailed PDF reports are also available for each buildings
rooftop. The user can input an address in order to locate a
given building whilst also receiving suggestions of similar
addresses from the server. Moreover, the shadowing caused by
vegetation and other buildings can be seen in Figure 7. The
4 Copyright (c) IARIA, 2014. ISBN: 978-1-61208-333-9
WEB 2014 : The Second International Conference on Building and Exploring Web Based Environments
Figure 7. Detailed overview of the calculated solar potential for the selected
building. The marked area in the red circle represents an example of reduced
solar potential due to shadowing from high-vegetation.
deciduous type of vegetation is less dense during the winter,
and has lower impact on the rooftop shadowing. In Figure 8,
the Android application is shown that is a light version of the
web-application.
Figure 8. Screenshot of the mobile application, where the used language is
Slovene. Two views are shown, namely the map view (left) and the details
view (right), where the graph of the monthly estimations is displayed.
There are no signicant bottlenecks on the server side,
since all the calculations are preprocessed beforehand. The
only possible time consuming operation is the generation of
the raster tiles that consist of multiple layers in the database.
These layers represent different data types: orthophoto, build-
ings shape layer and SP layer. Furthermore, the rasterization
operation of the geometry from SP layer is also one of the
heavier task for Geoserver. Due to possible server overload,
the users can notice slower loading times for tiles (especially
those with SP cells) in their client applications.
Additional statistics were made from the SP data. Detailed
overview on how the rooftops cells slope and aspect values
affects the received SP (see Figure 9). Total amount of grid
cells was 978282.
Figure 9. Graph of SP values through different cell (a) aspects and (b) slopes.
Figure 9.a represents the rooftop cell aspect (cell orienta-
tion), where the highest SP was observed at surfaces oriented
towards the geographic south (i.e., 0 aspect). On the second
graph (see 9.b), information on how the cell slope can affect
on the received SP. Cells with slope perpendicular to average
direction of the Suns radiation (i.e., near 45) can receive
higher irradiance than cells with other slopes.
V. CONCLUSION
In this paper, the development of a GIS application was
described that allows users to quickly locate suitable rooftop
areas with the help of the calculated solar potential. This
application provides deep insight into estimated monthly re-
ceived solar irradiance, as well as approximate prediction
of the produced electrical energy. Therefore, such type of
applications help improve sustainability of modern cities.
Furthermore, the presented application uses a recent method
for the solar potential estimation, where long-term diffuse and
direct irradiance measurements were considered, as well as
shadowing from vegetation and surrounding terrain. Addition-
ally, a mobile application was developed, where the user can
select the nearest rooftop by using the GPS technology. From
the implementation perspective, it is important to preprocess
as much of calculations as possible, in order to decrease the
computing load on the web servers. Additionally, the accuracy
5 Copyright (c) IARIA, 2014. ISBN: 978-1-61208-333-9
WEB 2014 : The Second International Conference on Building and Exploring Web Based Environments
highly depends on the input data (i.e., resolution of LiDAR
point cloud and long term irradiance measurements), therefore
when creating similar applications in the future, modern clas-
sication methods for LiDAR data should be considered, and
an accurate solar irradiance model.
VI. ACKNOWLEDGEMENTS
Thanks to the Surveying and Mapping Authority of the
Republic of Slovenia and Slovenian Environment Agency
for providing the estate and irradiance data. Additionally,
we would like to thank the Municipality of Beltinci that
was coordinating the regional project DoFPoLO, which was
supported by the Ministry of Education, Science, and Sport, as
well the European Regional Development Fund. This research
was supported by grants J2-5479 and P2-0041.
REFERENCES
[1] N. Luka c, D.

Zlaus, S. Seme, B.

Zalik, and G.

Stumberger, Rating
of roofs surfaces regarding their solar potential and suitability for PV
systems, based on LiDAR data, Applied Energy, vol. 102, 2012, pp.
803 812.
[2] N. Luka c and B.

Zalik, GPU-based roofs solar potential estimation
using LiDAR data, Computers & Geosciences, vol. 52, 2013, pp. 34
41.
[3] G. Petrie and C. Toth, Airborne and spaceborne laser prolers and
scanners, In: Shan J, Toth CK, editors. Airborne and spaceborne laser
prolers and scanners. Boca Raton: CRC Press, vol. 54, 2008, pp. 29
86.
[4] W. Hetrick, P. Rich, F. Barnes, and S. Weiss, GIS-based solar radiation
ux models. American Soc Photogrammetry & Remote Ssnsing+ Amer
Ccong ON., vol. 3, 1993, p. 132.
[5] R. Dubayah and P. Rich, Topographic solar radiation models for GIS,
International Journal of Geographical Information Systems, vol. 9(4),
1995, pp. 405419.
[6] L. Kumar, A. Skidmore, and E. Knowles, Modelling topographic
variation in solar radiation in a GIS environment, International Journal
of Geographical Information Science, vol. 11(5), 1997, pp. 475497.
[7] P. Fu and P. M. Rich., A geometric solar radiation model with
applications in agriculture and forestry, Computers and electronics in
agriculture, vol. 37(1), 2002, pp. 2535.
[8] J. G. Corripio, Vectorial algebra algorithms for calculating terrain
parameters from dems and solar radiation modelling in mountainous
terrain, International Journal of Geographical Information Science, vol.
17(1), 2003, pp. 123.
[9] M.

S uri and J. Hoerka, A new GIS-based solar radiation model and
its application to photovoltaic assessments, Transactions in GIS, vol. 8,
2004, pp. 175190.
[10] J. Ruiz-Arias, J. Tovar-Pescador, D. Pozo-V azquez, and H. Alsamamra,
A comparative analysis of DEM-based models to estimate the solar
radiation in mountainous terrain, International Journal of Geographical
Information Science, vol. 23(8), 2009, pp. 10491076.
[11] T. Voegtle, E. Steinle, and D. Tovari, Airborne Laser scanning data for
determination of suitable areas for photovoltaics, Remote Sensing and
Spatial Information Sciences, vol. 36(3/W19), 2005, pp. 215220.
[12] R. Kassner, W. Koppe, T. Sch uttenberg, and G. Bareth, Analysis of the
solar potential of roofs by using ofcial LiDAR data, In Proceedings
of the International Society for Photogrammetry, Remote Sensing and
Spatial Information Sciences, (ISPRS Congress), 2008, pp. 399404.
[13] B. Jochem, H oe, M. Hollaus, and M. Rutzinger, Object detection in
airborne LiDAR data for improved solar radiation modeling in urban
areas, Remote Sensing and Spatial Information Sciences, vol. 38(part
3), 2009, p. W8.
[14] J. Hoerka and J. Ka nuk., Assessment of photovoltaic potential in
urban areas using open-source solar radiation tools, Renewable Energy,
vol. 34(10), 2009, pp. 22062214.
[15] A. Jochem, H. Bernhard, and R. Martin, Extraction of vertical walls
from mobile laser scanning data for solar potential assessment, Remote
Sensing, vol. 3(4), 2011, pp. 650667.
[16] R. Levinson, H. Akbari, M. Pomerantz, and S. Gupta, Solar access of
residential rooftops in four California cities, Solar Energy, vol. 83(12),
2009, pp. 21202135.
[17] City of Anaheim Solar Map Template. [retrieved: 10, 2013]. [Online].
Available: http://anaheim.solarmap.org/
[18] Berkley Solar Map. [retrieved: 10, 2013]. [Online]. Available:
http://berkeley.solarmap.org/
[19] Boston Solar. [retrieved: 10, 2013]. [Online]. Available: http:
//gis.cityofboston.gov/SolarBoston
[20] DRCOG Denver Regional Solar Map. [retrieved: 10, 2013]. [Online].
Available: http://solarmap.drcog.org/
[21] LA Country Solar Tool and Green Planning Tool. [retrieved: 10, 2013].
[Online]. Available: http://solarmap.lacounty.gov/
[22] City of Madison. Wis MadiSUN. [retrieved: 10, 2013]. [Online].
Available: http://solarmap.cityofmadison.com/madisun/
[23] NYC Solar Map. [retrieved: 10, 2013]. [Online]. Available: http:
//nycsolarmap.com/
[24] Salt Lake City Solar. [retrieved: 10, 2013]. [Online]. Available:
http://www.slcgovsolar.com/
[25] San Diego Solar. [retrieved: 10, 2013]. [Online]. Available: http:
//sd.solarmap.org/solar/index.php
[26] San Francisco Solar Map. [retrieved: 10, 2013]. [Online]. Available:
http://sf.solarmap.org/
[27] Mapdwell Solar System. [retrieved: 10, 2013]. [Online]. Available:
http://www.cambridgema.gov/solar/
[28] R. C. F. Jakubiec, J. A., Towards validated urban photovoltaic potential
and solar radiation maps based on lidar measurements, GIS data, and
hourly daysim simulations, in 5th National Conference of IBPSA-USA,
Madison, Wisconsin., 2012.
[29] Dach acheneignung f ur Solaranlagen. [retrieved: 12, 2013]. [Online].
Available: http://geo.gkd-el.de/website/solar/viewer.htm
[30] Bristol. [retrieved: 12, 2013]. [Online]. Avail-
able: http://maps.bristol.gov.uk/pinpoint/?service=localinfo&maptype=
js&layer=Neighbouring+authorities;Solar+potential
[31] Wien Umweltgut. [retrieved: 11, 2013]. [Online]. Available: http:
//www.wien.gv.at/umweltgut/public/grak.aspx?ThemePage=9
[32] J. J. Michalsky, The Astronomical Almanacs algorithm for approxi-
mate solar position. Solar Energy, vol. 40.3, 1988, pp. 227235.
[33] R. C. F. Jakubiec, J. A., A method for predicting city-wide electricity
gains from photovoltaic panels based on LiDAR and GIS data combined
with hourly Daysim simulations, Solar Energy, vol. 93, 2013, pp. 127
143.
[34] N. Luka c, S. Seme, D.

Zlaus, G.

Stumberger, and B.

Zalik, Build-
ings roofs photovoltaic potential assessment based on LiDAR (Light
Detection And Ranging) data, Energy, vol. In Press, 2014.
[35] Quantum GIS. [retrieved: 10, 2013]. [Online]. Available: http:
//www.qgis.org/en/site/
[36] ILWIS. [retrieved: 10, 2013]. [Online]. Available: http://www.ilwis.org/
[37] GeoServer. [retrieved: 09, 2013]. [Online]. Available: http://geoserver.
org/display/GEOS/Welcome
[38] Mapnik. [retrieved: 09, 2013]. [Online]. Available: http://mapnik.org/
[39] Open Geospatial Consortion. [retrieved: 10, 2013]. [Online]. Available:
http://www.opengeospatial.org/standards
[40] PostgreSQL Database. [retrieved: 10, 2013]. [Online]. Available:
http://www.postgresql.org/
[41] PostGIS - Spatial and Geographic objects for PostgreSQL. [retrieved:
10, 2013]. [Online]. Available: http://postgis.net/
[42] OpenLayers. [retrieved: 10, 2013]. [Online]. Available: http:
//openlayers.org
[43] JQuery. [retrieved: 11.2013]. [Online]. Available: http://jquery.com/
[44] SolarEnergo. [retrieved: 03, 2014]. [Online]. Available: http://
solarenergo.beltinci.si/en/
6 Copyright (c) IARIA, 2014. ISBN: 978-1-61208-333-9
WEB 2014 : The Second International Conference on Building and Exploring Web Based Environments