Reverse AOD via Linke Turbidity ...

[Mark Mikofski]

... given or assuming pwat?

In fact perhaps given relative humidity or dew point and ambient temperature and pressure, couldn't we already reverse pwat? Then use TL to reverse AOD for use in simplified Solis or bird models?

Does anyone think this might be useful?

[Anton Driesse]

Hi Mark,

Maybe, but I'm not sure I understand the suggestion fully. Generally
though, I would rather use an end-to-end model, i.e. that uses the
parameters I have and gives the output I need, rather than using a chain
of models/transformations.

--
PV Performance Labs Germany
Emmy-Noether-Str. 2
79110 Freiburg
Germany

+49-761-8973-5603 (Office)
+49-174-532-7677 (Mobile)

www.pvperformancelabs.com

[Mark Mikofski]

Thanks Anton

My problem is that the Ineichen model has been causing issues, but it's really the only model that's ready to go with no extra data. I would like to recommend that people use the Simplified Solis model, but it requires AOD data to be provided separately, which most people don't have. But we do have this static TL table from 2013, and a way to calculate Pwat from Chris Gueymard. So I was thinking can't we just back out the AOD from there, and then use it in the Simplified Solis model? Problem solved, unless it's the Linke Turbidity data that's the problem. But Cliff and I looked at the TL data and compared it with clear days from the SURFRAD data as well as other models with AOD measurements/re-analysis from the McClear/CAMS model, andd the TL data seemed to stand up.

Anyway, in principle I agree with you, better to use an end-to-end model with the raw data. I'm not saying we should change any models. Just provide a way to generate AOD for any location from the static/historical TL data + local weather measurements like RH, temp, & press. If you had the AOD data already, then you wouldn't need to back it out, unless you wanted to calculate the Angstom alpha exponent.

Another option to this would be to include a static source of broadband (or 700nm) AOD data like we do now for TL.

Or another option would be to include more AOD retrieval tools in the iotools subpackage, in addition to the existing ECMWF which can be slow IMO.Maybe a way to retrieve Aeronet and other data sources if they're useful.

[Anton Driesse]

Hi Mark,

I was going to suggest developing a new clear sky model as an alternative in my previous post, but then I didn't.

What issues does Ineichen cause? 

It sounds like you want to use the same input (TL) but get something else out... which sound like a new model!

Anton

--
You received this message because you are subscribed to the Google Groups "pvlib-python" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pvlib-python...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/pvlib-python/1279302d-af34-4e8a-8e3f-ad3f958500f1%40googlegroups.com.

-- 
Photovoltaic Performance Labs Germany
Emmy-Noether-Str. 2
79110 Freiburg
Germany

[William Holmgren]

Ineichen's "Validation of models that estimate the clear sky global and beam solar irradiance" (2016) does a nice job of documenting the poor limiting behavior of a handful of clear sky models and the relatively good behavior of the Simplified SOLIS model. (attached)

Pvlib's Linke Turbidity climatology is convenient. I also don't trust it on scales smaller than a continent - that's why I added the climatology maps to the documentation: https://pvlib-python.readthedocs.io/en/stable/clearsky.html In my opinion, the fineness of the gridded data combined with the way pvlib's lookup functions work give a false sense of precision.

Someone could look into adding a gridded climatology of AOD and maybe PW for use with Simplified SOLIS.

Will

To view this discussion on the web visit https://groups.google.com/d/msgid/pvlib-python/09f0545e-1354-daa6-a2d6-e0ae75ca3821%40pvperformancelabs.com.

[Mark Mikofski]

Reviving this before ECMWF MACC model is deprecated forever. This is a source of AOD at many different wavelengths that could be used for Simplified Solis. We could download a global data set for 380-nm and 550-nm to use in Bird and to calculate broadband AOD for Simplified Solis. Any takers?