We are looking for a weather forecasting service (ideally free) that will provide hourly values for at least the next 24 hours. In addition to the temperatures, we need values for the solar radiation (aka irradiance), or possibly illuminance.
http://www.wunderground.com/ provides such a service and their first tier is free, but does not provide data for the solar radiation.
Is there such a service somewhere?
"National Solar Radiation Database (NSRDB) offers hourly solar radiation data including global, direct, and diffuse radiation data as well as meteorological data for those stations provided from NCDC’s Integrated Surface Database."
https://www.ncdc.noaa.gov/data-access/land-based-station-data/land-based-datasets/solar-radiation
Yes, I know this doesn't actually answer the question, but it was getting a bit long for a comment.
Most 'space weather' forecasting is concerned with erruptive geoeffective events (the liklihood of a flare, radio burst or a CME, and the chance of it having an effect here on earth). For space weather forecasts, the official US source is NOAA's Space Weather Prediction Center. NASA's Community Coordinated Modeling Center used to give forecasts for the liklihood of satellites and people in space being affected by events (where they don't have the same level of shielding from the earth's magnetic field), but there was an incident a few years back when NASA and NOAA gave conflicting forecasts, and now NASA doesn't give any at all.
Although there don't tend to be forecasts for irradiance, there are two major archives for it that I'm aware of -- besides NOAA's National Solar Radiation Database for ground based data that's already been mentioned, the University of Colorado Boulder's Laboratory for Atmospheric and Space Physics's LASP Interactive Solar Irradiance Datacenter has irradiance measurements from spacecraft.
One huge problem with irradiance data by location is much of the same problem with day length by location -- you have to take into consideration things like obstructions (mountains, buildings, eclipses) and other effects of geography (day length will be longer, thus irradiance greater if on the equator side of a mountain in the summer). If there are any irradiance forecasts, I would expect them to be non-location specific.
Although the CCMC no longer provides forecasts, they do collect up an awful lot of near-real-time data at the Integrated Space Weather Analysis System
(disclaimer : I'm attached to multiple solar physics projects, one of which provides irradiance data to these groups)
pvliblibrary but I would prefer 1-hr granularity of my forecasts going out beyond 7-9 days.