All right. If we can get people seated. All right. Can you hear me? Yeah. I hope you got your coffee because we have an exciting last session of the day here about a little

bit of everything. And the first talk is going to be the solar radiation model for open source GIS implementation and applications. And Marcel Suri is going to give the talk. And he wants the mic. Thanks. Good afternoon. I'm coming from a company

for geomodel. Now I'm also working in the European Commission Joint Research Center where I am involved in a project where the primary task was to build a solar radiation database at different scales in Europe. The problem was that to build such databases,

we didn't find the tools which could be helpful for us. So this forced us to develop a new solar radiation model which overcame problems which were found in other solutions. Here

I wanted also to say that in our company we use not only open source GIS software but also commercial software. And we have found quite a good combination in using both

systems which helped us to overcome different problems. But the development of the model was done in GRAS because GRAS is a really powerful system for building models with

good possibility to run long computations in batch processing. In this presentation I will get first some comments on the solar radiation, the problems of its modeling,

the methods which can be used for deriving the GIS maps. Then I will have some comments on the solar model itself. And I would like to present some applications.

When talking about solar radiation, this is quite interesting in many applications, technical, environmental, land use planning and so on. The problem with solar radiation

is that it is measured on a quite limited number of metallurgical stations around the world. The situation is better in Europe but still not satisfactory to derive GIS maps.

Therefore, as you can see on this example of Central and Eastern Europe, from this relatively limited number of stations we have to provide quite high resolution maps.

So the primary question is what is the solar radiation available for our application in a site which is in relatively large distance from the neighboring stations.

There are three groups of methods which are used to derive spatially distributed maps. Spatial interpolation techniques are relatively easy but because solar radiation is determined by a complex set of factors, these methods are usually not satisfactory for providing

results in regional scales. Therefore, multivariate methods are used which take into account also relief or satellite data or other contextual information.

The second group of methods which is found to be relatively quite successful is processing of satellite data, I mean metallurgical data coming from geostationary satellites like Meteosat and so on. The third approach which is highly promising

and you will see the results is GIS-based modeling. Certainly, the most useful approach is to combine two or three methods as described here.

When talking about solar radiation modeling, I have to say that there are factors which are relatively easy to model, which can have quite clear physical behavior.

First of all, there are factors connected with the Earth's rotation revolution and relief or terrain where the problems with modeling don't exist.

Then, the third group of factors is atmospheric attenuation where the uncertainty in deriving the necessary spatial data is growing from top to bottom. The most problematic factor to be modeled are clouds and also solid and liquid particles

which are available in atmosphere. So now, how the model works. Here are the listed input parameters

which are necessary for modeling solar radiation. The first group is necessary to model so-called clear sky radiation

when we don't consider the effect of clouds. So, we consider clear atmosphere, sunny weather situation. The difference between irradiance and irradiation is that the irradiance gives us

the values of the solar power which is available at the time moment while irradiation represents solar energy which is available in a certain time interval

for example one day, one month or so. The red marked parameters which are used to the model mean the so-called problematic factors. Linked turbidity which expresses the influence of the solid particles in the atmosphere

cannot be described by deterministic models and usually this parameter is derived using empirical estimations. There are some possibilities where to derive the values in our PDF presentation.

You can find also some internet links. To model real sky irradiance or irradiation we need information about the status of cloudiness.

For averages like monthly averages or annual averages quite often used empirical parameter is used which is called clear sky index which describe the attenuation of clear sky atmosphere

by all the components of the radiation of beam and diffuse. I don't want to go into big details. So, what are the outputs of the model? First of all, there are three components of the solar radiation which are computed.

As you can see there are two modes. You can run the model just for one time moment, one second or you can run it on a basis of calculating the sums for one day, one month or so.

Then there are some other parameters which can be calculated like solar incidence angle or duration of beam irradiation during one day. Some data outputs are also written to text file.

Again, just to remark that the real sky modeling or the modeling of real sky conditions depends on the availability of information about cloudiness which in our case was described by this empirical index, clear sky index.

In the first application, let me present the model outputs on the continental scale. You can see the Central and East European countries.

The solar database was built with aim to be used in the assessment of the performance of photovoltaic systems. Photovoltaic systems are solar panels as you can see here on this roof of the house

together with some other electrical equipment which are converting the sun energy to the electricity. In our project, the interest was focused to the systems which are directly connected to the electrical grid

to feed electricity to the grid during the day hours and vice versa to take electricity from the grid

for inhabitants of the house during the night hours. The tools we have used were first of all solar radiation model and then we have used the interpolation methods in GRASS,

two-dimensional interpolation methods, spline with tension and also 3D spline as well as RST. The data used were based on the digital terrain model of one kilometer grid resolution

and then we have used climatic data from 182 stations. The resulting database consists of monthly and annual averages of daily totals of global irradiation for solar panels at different positions.

The methodology looks like this. Here are the main input parameters, elevation, turbidity, latitude which were used to calculate so-called clear sky values

that means considering atmospheric without clouds attenuation. Then clouds effect were calculated from the metrological measurements. Here is the spatial interpolation using the multi-dimensional method,

which takes into account not only these solar measurements but also terrain. This is the result for real conditions.

Now let me give you some more detailed insight into the database. This is the solar radiation, global radiation, which means the sum of beam diffuse and reflected components

for solar panels which are inclined southwards at different inclination angles. This is quite important thing which was asked by engineers and technicians

which are configurating these photovoltaic systems because inclination angle certainly determines the amount of electricity which could the system provide. So you can see that the higher inclination angle,

the more electricity or the more solar energy is available for solar panels. But this is highly dependent on the latitude and here you can see that increasing the solar panel from horizontal position to the 15 degrees,

you can have up to 500 watts per day in average. Here is the increase of the energy income when considering these two angles.

But here you can see that increasing inclination of solar panels up to 40 degrees does not mean a big increase in solar energy input and in some areas, mostly below the latitude of 45 degrees,

there is an energy decrease. So that means that it is not meaningful to put solar panels higher than something about 25 degrees.

Here you can see the seasonal changes of solar radiation during the year which are relatively big. One of the most very important thing when talking about solar radiation database

is the so-called diffuse component which has to be considered in many applications. So you could see the time series. Here I just wanted to show that our results can be compared to the previously done.

Database is computed from using different approaches and the new features were found mostly because of better representation of terrain.

The second application is from Slovakia and it just shows the possibility of the model application for computing solar irradiance, that means the time in one radiation that is available at a certain moment of time.

Here you can see what is the difference between the beam, diffuse and reflected components. In many works, reflected components is just not considered because it adds to the global radiation only very small values.

Here is a series of two different hours computed for the spring equinox, difference between morning and noon values.

Here this example shows the difference of considering or not considering the so-called shadowing effect of the mountainous ridges

which can contribute quite significantly to lowering of radiation values. This is also another example where you can see the same methodological approach

applied on a dataset with a resolution of 1 km and 100 m. You can see that considering the shadowing effect has revealed quite significant lowering

of available solar radiation in January in mountainous areas which what you cannot see on this data, 1 km data. Well, because the database consists of time series, you can use also animations

to better communicate the temporal spatial and temporal pattern of the changing of the solar radiation on this extent. Then, just a few words to conclude, the new model is now available in the 5 versions.

It is complex and flexible and it is based on the latest European solar radiation research

which was published in different books and papers. And because of open source, it can be modified for any other purposes or after new improvements in the solar radiation research, the equations can be modified.

Thank you. One very quick question. That was not a quick one. Quick question?

Quick question. I didn't fully understand, maybe I didn't get the point. How is the cloud cover modelled? Is that just a factor which you apply or is it a localized map which you use?

The cloud cover was calculated from the ground measured values and the spatial database was calculated using the interpolation by S-WOW RST,

multi-dimensional interpolation. I think that more details you can find in our paper. It's not so easy to tell it in one sentence. Okay, we better move on.