gendaylit in DIVA 2.0 seriously underesitmates total irradiation?

Hi, DIVA list, I found that the gendaylit come with DIVA2.0 seems to seriously underestimate global irradiation.

For example, in Singapore's epw weather file, for Jan 1 9am:

global horizontal irradiation is 152 Wh/m^2

direct normal irradiation is 70 Wh/m^2

diffuse horizontal irradiation is 130 Wh/m^2

Use the gendayilt in C:\DIVA\DaysimBinaries (51KB in size), I can only get 94.07851 Wh/m^2 for an unobstructed upward sensor, which is far lower than the 152 Wh/m^2 as recorded in the weather file and seems to be way off the line... (script is shown in the end of this post). This underestimation seems to be consistent for whatever time, day and city specified.

However, I have an older version of gendaylit (47KB in size) which can produce an unobstructed global horizontal irradiation of 157.5287 Wh/m^2 which is quite close to the value recorded in the weather file. (this older version of gendaylit might come from Radiance 3.9 package)

I'd like to ask:

1) whether the gendaylit.exe come with DIVA2.0 has been modified and recompiled, and what the changes are?

2) why there's such a huge difference in the irradiation as predicted?

(the two different versions of the gendaylit.exe are attached for your reference)

Thanks!

-Ji

1) the content of the ./sky.rad file:

!gendaylit  1 1 9 -a 1.37 -o -103.98 -m -120 -W 70.0 130.0 -O 1
skyfunc glow sky_glow
0
0
4 1 1 1 0
sky_glow source sky
0
0
4 0 0 1 180


2) the content of the autorun.bat file

oconv ./sky.rad > ./sky.oct
echo 0 0 0 0 0 1 | rtrace -w -h -I+ -ab 1 ./sky.oct | rcalc -e "$1=$1*0.265+$2*0.670+$3*0.065" > ./results.txt

Views: 1162

Reply to This

Replies to This Discussion

Hi Ji, 

 

Thank you for the detailed post. Unfortunately, I cannot download the files you uploaded -- this seems to be a bug in the Ning site which we need to reconcile at some point. Is it possible to email me the files or post them on some other website?

 

Thanks so much,

Alstan 

Hi, Alstan, Thanks for your prompt reply!

Please download the new (come with DIVA2.0) and the old gendaylit.exe program I tested from the following links:

1. new version:

https://sites.google.com/site/episodehopezh/files/gendaylit_diva.ra...

2. old version:

https://sites.google.com/site/episodehopezh/files/gendaylit_old.rar...

each of the zip files contains:

1. the new or old version of gendaylit.exe

2. the sky file, the oct file, an hdr image of the sky,  and the global horizontal irradiation value for one unobstructed upward sensor for respective version of gendaylit.exe

Thanks again for you kind help!

- Cheers, Ji

FYI:

the hdr image of the sky generated by the gendaylit.exe come with DIVA2.0:

the hdr image of the sky generated by the old gendaylit.exe (47KB in size):

Indeed Ji, it looks like the Solar Source is being created properly but there is some error with the diffuse component. I'll be forwarding this along to the DAYSIM developers.

 

By the way, I attach my own similar analysis of the images.

 

Best,

Alstan 

 

 

Great! Thanks, Alstan! I'm sure your colleagues can help to take a look and solve this problem! - Cheers, Ji

BTW, Hi, Alstan, can you kindly share the script to generate the two falsecolor images you produced? I still don't seem to have a clue on how to manipulate falsecolor.exe or falsecolor2.exe... Thanks! - Ji

J. Alstan Jakubiec said:

Indeed Ji, it looks like the Solar Source is being created properly but there is some error with the diffuse component. I'll be forwarding this along to the DAYSIM developers.

 

By the way, I attach my own similar analysis of the images.

 

Best,

Alstan 

 

 

I just generated them in wxFalsecolor. You can see the options I used if you enlarge the screenshots.

 

There are a series of parameters you can pass to falsecolor2, which we implement in DIVA something like this:

falsecolor2 -i input.pic -s auto -n 10 -l kWhm-2 -z > output.pic

 

Alstan 

Understand, Alstan, much appreciated!

- Ji



J. Alstan Jakubiec said:

I just generated them in wxFalsecolor. You can see the options I used if you enlarge the screenshots.

 

There are a series of parameters you can pass to falsecolor2, which we implement in DIVA something like this:

falsecolor2 -i input.pic -s auto -n 10 -l kWhm-2 -z > output.pic

 

Alstan 

some follow up ... 

the falsecolor image (stereographic view type for rpict) of the sky irradiance produced by gendaylit and projected to a virtual sky hemisphere (y axis is due North) for 1st Jan 9am in Singapore using the Houdini software

(view with Tregenza 145-sky-patch subdivision)

Dear Ji,

Jan Wienold located  the source of the error. The WINDOWS version of gendaylit has problems reading in selected data files from the Radiance library. In order to remedy the problem immediately, just go to c:DIVA\Radiance\lib, open  coeff_perez.dat and defangle.dat and delete the three header lies starting with '#'. Our next DIVA update, version >2.0.0.7, will have this bug corrected. Thank you for pointing this out.

Best,

Christoph

Noted with many thanks, Christoph and Jan!

- Cheers, Ji

For the sake of closure, we just uploaded a new installer that has a version of GenDaylit which performs as expected.

 

All the best,
Alstan 

Reply to Discussion

RSS

© 2017   Created by jeff niemasz.   Powered by

Badges  |  Report an Issue  |  Terms of Service