Spectral index doesn't change when converting between temperature and flux density

[jpober]

Temperature and flux density should, in principle, have different spectral indices. (In Rayleigh-Jeans limit, a blackbody has spectral index 0 in temperature and spectral index +2 in flux density.) We don't change our spectral index parameter when we change the units of stokes. We should consider doing that, or at least warn.