Dear Radiance users,
I am sharing some of our recent work that may be of interest to those in the photovoltaics (PV) sector, or anyone running thermal studies that require spectrally-resolved simulations beyond the visible band.
We’ve developed a Python-based sky generator that is essentially a spectral extension of gendaylit — therefore the name gensdaylit (the “s” is for spectral). Simulations proceed as usual; you simply create the sky .rad file with gensdaylit.
I understand that there are other alternatives, but in PV applications gendaylit is the preferred method as it is based on the commonly used Perez All Weather sky model.
gensdaylit characterizes the sun and sky spectra across the full solar spectrum using Radiance’s multispectral modifiers. To compute the sun and sky spectra we used SMARTS coupled with site-specific aerosol optical depth and precipitable water vapor data from AERONET. Of course, you can use any other atmospheric radiative transfer model (e.g., libRadtran).
In addition, gensdaylit integrates a solar-disk sampler for efficient penumbra rendering.
You can read more in the recently published paper, or experiment with the code directly.
A validation is underway with promising results, using a leaf-resolved tree canopy model in a complex scene. I’ll share it once published.
Feel free to reach out, here, or on LinkedIn.
This would not have been possible without this forum. Thank you, especially to Greg and Taoning.
Cheers,
Alex
P.S. Radiance is a wild beast and I make no claim to expertise, so apologies if I’ve missed something.