File talk:Solar spectrum en.svg
The “correction” from “W/m²/nm” to “W/m²” is, I'm afraid, wrong. The first is correct - the height of the graph is the amount of power arriving in the area for a given part of the spectrum. For example, roughly 1.4 W/m² arrive at the surface in the range 500 to 501 nm and the same again in the range 501 to 502 nm. Obviously, reality is a bit more continuous but that's the idea.
Formally, following the SI style guide, the units should be GW/m²/m (or GW/m³) as you're supposed to use base units except for the initial prefix. However, that'd be rather confusing here and I wouldn't recommend it. EdDavies (talk) 08:17, 4 October 2014 (UTC)
Agree with the above comment. You should arrive at "W/m^2" by integrating over a part of the curve or the whole curve. This means, you have to start with "W/m^2/nm" on the y axis, and "nm" on the x axis.
Source data
edit
Is there a citation for the data used to composite this graphic? This could be useful to some readers wishing to either validate its accuracy or make use of the data. Tar-Elessar (talk) 19:09, 24 April 2018 (UTC)
W vs kW
edit
The latest change (14:13, 25 February 2019) of W to kW by Tuvalkin appears wrong. Not sure where the discussion referenced in the comment is, but if you look at the spectra like e.g. ASTM G-173, you'll see that irradiance at 452 nm is ~2 W/m²/nm, not kW/m²/nm. — Preceding unsigned comment added by L0ll0lll (talk • contribs)
