Question-and-Answer Resource for the Building Energy Modeling Community
 | 1 | initial version |
I assume you're referring to a location in North America, because of your mention of the NSRDB that covers only the Western Hemisphere from 60N down to 20S latitude. The solar radiation data in the new NSRDB (Sept. 2016) was derived from weather satellite observations, Prior to its release, all the solar data that you see on any weather file are modeled using various empirical and physical models based on the observed Cloud Cover, and other meteorological parameters.
If you're asking for satellite-derived solar later than 2015, these are available commercially but they' cost hundreds of dollars per site per year. Otherwise, you have to go to modeled solar, which nobody complained about until the satellite-derived solar became available :-)
I've been processing the ISD for over ten years, and have spent the majority of effort on modeling the solar radiation and spot checks against measured solar show that I'm within 5% on an annual basis. E-mail me if you want more details.
I'm not quite sure the point of your last question. If you're asking whether getting the right GHI/DNI then I would say that's very important. If you're asking whether it's okay to overlay historical weather data with GHI/DNI from another time or location, I would say don't do it because you would defeat the whole point of using historical weather data.
