subtended sun angle in gensky and gendaylit

radiance-general
1

Hi Francesco,
dear folk,

I have been in some discussions with a user about gendaylit and especially the cygwin- binary of it.
We found out, that the subtended angle of the sun differs from the original gendaylit (0.533) version to the one, which we can download on http://www.bozzograo.net/radiance(0.5 for cygwin binary).
For that reason, I checked in one of the main references in the solar world (Duffie, Beckman: Solar engineering of thermal processes) and found 32' for the subtended solar angle, which corresponds to 0.5333.

So I assume you changed the value in the source code, or? But as long as you are not adapting the radiance value of the light source (which you don't do), you'll loose more than 10% of the flux of the sun by reducing the value from 0.533 to 0.5 !!

Is there any reason to use 0.5 instead of 0.533?

And a question to Greg: gensky is also using 0.5. Is there any reference to that? But: As long as the overall flux is correct, the difference in the angle can be neglected.

For the cygwin-version of gendaylit I would suggest to correct that value in your version to 0.533 of the source code. In addition to this, I'll try to put a gendaylit version on our Web-page, which can be compiled also under cygwin without any modification. If ready, I'll announce this here again.

Cheers,

Jan

···

--
Dipl.-Ing. Jan Wienold
Project Manager
Fraunhofer-Institut f�r Solare Energiesysteme
Thermal Systems and Buildings, Lighting and Daylighting
Heidenhofstr. 2, 79110 Freiburg, Germany
Phone: +49(0)761 4588 5133 Fax:+49(0)761 4588 9133
[email protected]

In office: Mo,Tue: 9:00-18:00
We-Fr: 8:30-14:00

2

Hi Jan,

The 0.5 degree value in gensky is approximate, and could be corrected to 0.533. This is not difficult to do. The current radiance is figured using the 0.5 degree value, and could be adjusted. I see no reason not to make such a correction in the next release.

Other opinions?
-Greg

···

From: Jan Wienold <[email protected]>
Date: April 4, 2006 4:52:00 AM PDT

Hi Francesco,
dear folk,

I have been in some discussions with a user about gendaylit and especially the cygwin- binary of it.
We found out, that the subtended angle of the sun differs from the original gendaylit (0.533) version to the one, which we can download on http://www.bozzograo.net/radiance\(0.5 for cygwin binary).
For that reason, I checked in one of the main references in the solar world (Duffie, Beckman: Solar engineering of thermal processes) and found 32' for the subtended solar angle, which corresponds to 0.5333.

So I assume you changed the value in the source code, or? But as long as you are not adapting the radiance value of the light source (which you don't do), you'll loose more than 10% of the flux of the sun by reducing the value from 0.533 to 0.5 !!

Is there any reason to use 0.5 instead of 0.533?

And a question to Greg: gensky is also using 0.5. Is there any reference to that? But: As long as the overall flux is correct, the difference in the angle can be neglected.

For the cygwin-version of gendaylit I would suggest to correct that value in your version to 0.533 of the source code. In addition to this, I'll try to put a gendaylit version on our Web-page, which can be compiled also under cygwin without any modification. If ready, I'll announce this here again.

Cheers,

Jan

3

Greg/Jan,

> So I assume you changed the value in the source code, or? But as
> long as you are not adapting the radiance value of the light source
> (which you don't do), you'll loose more than 10% of the flux of the
> sun by reducing the value from 0.533 to 0.5 !!
>
> Is there any reason to use 0.5 instead of 0.533?

Whether this matters depends on how you determine the solar radiance. Usually, it is derived from measurements of direct normal irradiance or illuminance (commonly found in climate files). Rarely is the solar radiance or luminance measured directly. Provided the solar radiance is determined in a consistent fashion from measurements of direct normal, then there is negligible practical difference as far any evaluation is concerned - the flux will be the same whatever the solid angle used for the sun.

In any case, the best measurements of direct normal are taken using a tracking device that commonly has an acceptance angle of 6deg (with, I recall, a fairly flat response). This is because it is impractical to attempt to track the sun using anything with a much smaller angle, let alone an angle that matched exactly the solar disc. So, when using these data from climate files etc., the difference between 0.5 and 0.533 is made pretty much irrelevant because the sun radiance is being calculated from a measurement that includes a lot of circumsolar region also.

Intriguingly, in practical terms, this matters more for overcast rather than clear sky conditions (I can almost hear the gasps of disbelief). For clear skies, the solar radiance dominates the brightness of of the circumsolar region -- so the "contamination" that results from basing the solar radiance on a measurement that includes the contribution of the circumsolar region is small. However, when a sky is overcast, the measurement of direct normal (in the climate file) is that which results from a (sunless) 6deg patch of sky. Say that value is used routinely in an annual calculation procedure to determine the radiance of the 0.5deg sun. Then, for overcast days, the solar radiance will be about ~144 times the radiance of the background sky (i.e. [6/0.5]**2). So, a brightish sun will be a permanent feature in all your overcast skies. If your evaluation is very sensitive to the magnitude of the sky and sun radiance or luminance (as I suspect Jan's is), then this may be an issue. It's probably wise to play around with a threshold value below which the sun radiance is set to zero.

-John

PS. There is some discussion on related matters in chapter 3 here:
http://www.iesd.dmu.ac.uk/~jm/zxcv-thesis/

···

-----------------------------------------------
Dr. John Mardaljevic
Senior Research Fellow
Institute of Energy and Sustainable Development
De Montfort University
The Gateway
Leicester
LE1 9BH, UK
+44 (0) 116 257 7972
+44 (0) 116 257 7981 (fax)

[email protected]
http://www.iesd.dmu.ac.uk/~jm

4

Hi Jan

So I assume you changed the value in the source code, or? But as long as
you are not adapting the radiance value of the light source (which you
don't do), you'll loose more than 10% of the flux of the sun by reducing
the value from 0.533 to 0.5 !!

I've just noticed that the cygwin and mingw versions of gendaylit
produce different outputs. I think that I might have used a different
and older source package when I compiled the first cygwin version,
which also doesn't include the -G option.
At least I've understood that I didn't change the gendaylit code :stuck_out_tongue:

I also agree with John about the weather data measurement accuracy,
and usually totally discard the direct normal solar component if lower
than 120 W/m2, as recommended by the WMO. I think S@tel-Light uses
a similar threshold to decide whether there is direct sun or not.

Ciao,

Francesco

5

Dear John,

I meant a different thing: The implementation of the Perez sky model in RADIANCE is made in a way, that you plug in site information and direct and diffuse irradiation (regardless if those values a measured data or test reference years) and you will get a RADIANCE description of the sun as light and source. This model has been validated on different datasets in the past. If you change then the size of the source(from 0.533 to 0.5) without a change of the "light" parameters (which are produce by gendaylit) and keeping all other factors, than you decrease the total flux of the sun (not of the sky, this remains the same since the parameters are the same). And therefore you get different results than using the original version.

This has nothing to do with the accuracy of the input radiation data, which is another issue which could be discussed.
It's just a simple implementation matter of the tool. I totally agree that the size change of the sun has no significant impact as long as the total flux is kept (which isn't in our case here).

Jan

John Mardaljevic wrote:

···

Greg/Jan,

> So I assume you changed the value in the source code, or? But as
> long as you are not adapting the radiance value of the light source
> (which you don't do), you'll loose more than 10% of the flux of the
> sun by reducing the value from 0.533 to 0.5 !!
>
> Is there any reason to use 0.5 instead of 0.533?

Whether this matters depends on how you determine the solar radiance. Usually, it is derived from measurements of direct normal irradiance or illuminance (commonly found in climate files). Rarely is the solar radiance or luminance measured directly. Provided the solar radiance is determined in a consistent fashion from measurements of direct normal, then there is negligible practical difference as far any evaluation is concerned - the flux will be the same whatever the solid angle used for the sun.

In any case, the best measurements of direct normal are taken using a tracking device that commonly has an acceptance angle of 6deg (with, I recall, a fairly flat response). This is because it is impractical to attempt to track the sun using anything with a much smaller angle, let alone an angle that matched exactly the solar disc. So, when using these data from climate files etc., the difference between 0.5 and 0.533 is made pretty much irrelevant because the sun radiance is being calculated from a measurement that includes a lot of circumsolar region also.

Intriguingly, in practical terms, this matters more for overcast rather than clear sky conditions (I can almost hear the gasps of disbelief). For clear skies, the solar radiance dominates the brightness of of the circumsolar region -- so the "contamination" that results from basing the solar radiance on a measurement that includes the contribution of the circumsolar region is small. However, when a sky is overcast, the measurement of direct normal (in the climate file) is that which results from a (sunless) 6deg patch of sky. Say that value is used routinely in an annual calculation procedure to determine the radiance of the 0.5deg sun. Then, for overcast days, the solar radiance will be about ~144 times the radiance of the background sky (i.e. [6/0.5]**2). So, a brightish sun will be a permanent feature in all your overcast skies. If your evaluation is very sensitive to the magnitude of the sky and sun radiance or luminance (as I suspect Jan's is), then this may be an issue. It's probably wise to play around with a threshold value below which the sun radiance is set to zero.

-John

PS. There is some discussion on related matters in chapter 3 here:
Institute of Energy and Sustainable Development

-----------------------------------------------
Dr. John Mardaljevic
Senior Research Fellow
Institute of Energy and Sustainable Development
De Montfort University
The Gateway
Leicester
LE1 9BH, UK
+44 (0) 116 257 7972
+44 (0) 116 257 7981 (fax)

[email protected]
Institute of Energy and Sustainable Development

_______________________________________________
Radiance-general mailing list
[email protected]
http://www.radiance-online.org/mailman/listinfo/radiance-general

--
Dipl.-Ing. Jan Wienold
Project Manager
Fraunhofer-Institut f�r Solare Energiesysteme
Thermal Systems and Buildings, Lighting and Daylighting
Heidenhofstr. 2, 79110 Freiburg, Germany
Phone: +49(0)761 4588 5133 Fax:+49(0)761 4588 9133
[email protected]

In office: Mo,Tue: 9:00-18:00
We-Fr: 8:30-14:00

6

Hi Jan,

I followed your discussion in the first e-mail. Reading it did remind me of the (sort of) related issue which I came across a while back and always meant to share with this group, but never got round to doing so before. Hence my contribution.

As this thread has been broadened somewhat (or slightly hijacked): Christoph - do you set a direct normal threshold in DAYSIM as suggested by Francesco?

> and usually totally discard the direct normal solar component if lower
> than 120 W/m2

-John

···

-----------------------------------------------
Dr. John Mardaljevic
Senior Research Fellow
Institute of Energy and Sustainable Development
De Montfort University
The Gateway
Leicester
LE1 9BH, UK
+44 (0) 116 257 7972
+44 (0) 116 257 7981 (fax)

[email protected]
http://www.iesd.dmu.ac.uk/~jm