Hello,

There are several MeteoSuisse weather stations in the canton of Zurich:
Kloten (airport), Afflotern, Fluntern, etc. (http://www.meteosuisse.admin.
ch/home/systemes-de-mesure-et-de-prevision/stations-au-sol/
reseau-de-mesures-automatiques.html?station=klo) . There is also a
commercial provider of TMY-style files called METEONORM (www.meteonorm.ch),
which I happily used for about 4 years. They can do typical files output in
a variety of formats. Generally, their files based on newer data (esp. for
Europe) tend to be more extreme than the IWEC files (or other EPW files
available on the EnergyPlus website).

To add to what Joe said about using models for splitting GHI into DNI and
DHI: I think the "Perez split" that the Big Ladder folks are referring to
is from Perez's 1992 paper (DYNAMIC GLOBAL-TO-DIRECT IRRADIANCE CONVERSION
MODELS, ASHRAE RP-644) . I've come across this problem before myself, and
have generally chosen one of the models coded in the PVPMC modelling
toolbox: https://pvpmc.sandia.gov/applications/pv_lib-toolbox/ . They have
several models available: Erbs, Louche, Orgill and Hollands, etc. They
don't recommend one or another and neither would I. I think they're all
quite bad for individual predictions, but might be OK to reproduce summary
statistics.

Best,

Parag

On 13 November 2016 at 06:01, via Bldg-sim
wrote: