It is well known that the surface temperature of the sun is determined by fitting the sunlight with the black-body spectrum.
Why is this inappropriate for Venus?
I have the question because of the History SE question When did we figure out that Venus was too hot for humans?1. It seems that it is impossible to determine the surface temperature of Venus as we did for the Sun on earth.
1While it looks like it should have been asked on HSM, it predates that site.
The important difference is that the word "surface" has a different meaning depending on whether we're talking about the Sun or Venus.
When we say "the surface of the Sun," we essentially mean "the visible part of the Sun." It's that part of the sun that emits light that we see, as opposed to the part of the sun where light gets emitted but then re-absorbed.
On the other hand, when we say "the surface of Venus," we certainly don't mean "the visible part of Venus." We mean the surface of the rocky part of Venus—the place where the gas ends and the solid begins. That part is, of course, hidden beneath Venus's thick atmosphere.
So basically, the reason we can see the Sun's surface but we can't see Venus's surface is that the phrase "the Sun's surface" is defined as meaning "the part of the Sun we can see," but the phrase "Venus's surface" is not defined as meaning "the part of Venus we can see."
I think the problem is that spectroscopy at visible and infrared wavelengths simply can't penetrate the thick atmosphere of Venus. Thus the information comes from much higher in the atmosphere and not from the surface.
The Sun doesn't have a surface. However, it does have a reasonably sharply defined photosphere, which is the "surface" from which the light reaching us escapes. The "effective temperature" of the photosphere is then just the temperature that a blackbody of that radius would have to have in order to emit the luminosity of the Sun.
Thus the effective temperature isn't actually the temperature of anything, it is a parameter relating the radius of the Sun to its luminous output. In fact the solar spectrum isn't a blackbody spectrum and its peak combined with Wien's law wouldn't yield the effective temperature.
All this is by way of explaining the difficulties of Venus. The planet will also have a "photosphere" from which light of any particular wavelength will escape. For light at optical and infrared wavelengths this will be well above the surface, and modelling that spectrum will give a temperature way lower than the surface temperature.
In order to get the surface temperature one would need the temperature at this Venusian photosphere, an idea of how high above the surface that was, and some knowledge of the temperature gradient of the atmosphere.
All this can be worked out with an appropriate mixture of physical models and observations at different wavelengths that probe to different depths in the atmosphere.
