Repository navigation
Attributing circumsolar irradiance to diffuse or direct POA #2874
Description
Transposition models split the diffuse irradiance into several components, including circumsolar, isotropic and horizon brightening. The main output of a transposition model is the irradiance on a tilted surace, poa_global, and its component, including poa_direct, poa_diffuse, poa_sky_diffuse, poa_ground_diffuse.
pvlib's general purpose transposition function (get_total_irradiance) attibutes the circumsolar irradiance in poa_sky_diffuse.
In contrast, pvlib's two view factor models infinite sheds and ANTS-2D attribute the circumsolar irradiance to poa_direct.
The argument for including circumsolar in poa_sky_diffuse is that it is measured diffuse irradiance. The argument for including circumsolar in poa_direct is that the irradiance originates from around the sun disk and thus is treated as direct irradiance when transposing irradiance or estimating angle of incidence losses.
Activity
cwhanse commentedon Sep 29, 2026
Some external context:
From PVsyst https://www.pvsyst.com/help/physical-models-used/irradiation-models/transposition-model.html#accuracy-and-validations:
I don't know what the "new treatment" is.
markcampanelli commentedon Sep 29, 2026
I think this issue was mentioned here: #2801 (comment)
"...we should try to indicate when, say, the circumsolar or horizon components are considered "concentrated", because that may affect how downstream consumers compute with the provided values. Including this info here seems to me like a stretch, but considering the downstream consumers NOW might shed some more light on if this would be worth the additional complexity."
adriesse commentedon Sep 30, 2026
The reality of circumsolar diffuse is that it's neither isotropic nor concentrated near the sun, so which way you lean can rightly depend on the application.
cwhanse commentedon Sep 30, 2026
Can you elaborate on the second point? I thought circumsolar was concentrated in an annulus around the sun.
adriesse commentedon Sep 30, 2026
Yes, around the sun, but not necessarily all near the sun as shown in this simulated sky: