Blackbody or characteristic emission of radiation

Question

I am trying to understand what determines to what degree energy from a "hot" object is emitted as characteristic radiation or blackbody radiation.

For example in a gas discharge lamp, a considerable amount of the energy emitted by the hot gas is clearly emitted as characteristic radiation, but some of it must also be emitted as blackbody radiation.

Another example would be a gas, such as a greenhouse gas, e.g. CO2, in the atmosphere absorbing, terrestrial radiation. When discussing the greenhouse effect, it is often simply stated that this energy is emitted as blackbody radiation by the atmosphere. Is emission of characteristic radiation according to allowed transitions not relevant?

UPDATE

*By characteristic radiation I mean line emission i.e. emission due to electronic, rotational, vibronic... transitions.

*The question was prompted by a project on the greenhouse effect, so I am mainly interested in whether it is accurate to say that the energy absorbed by the atmosphere from the thermal radiation of the earth, is subsequently emitted as blackbody radiation. John Rennie's answer seems to suggest this is not the case. If so, why is the greenhouse model explained in terms of the atmosphere doing precisely that?

Answer 1

The dominant source of black body radiation are transient oscillating dipoles induced by thermal vibrations within the material.

If we treat a solid as a cloud of electrons intermingled with a cloud of nuclei, then any thermally induced vibrations will cause small local changes in the average electron and nucleus density, and this will result in a small local electric dipole. As these dipoles change with time they emit the electromagnetic radiation that we see as black body radiation.

This applies to any system dense enough to support these sorts of charge density oscillations. Under normal conditions this means a solid or a liquid, but a plasma will behave in the same way if it can be made dense enough. This is hard to do in the laboratory but the Sun manages it :-)

However a gas cannot sustain oscillations of this form because the gas molecules spend most of their time distant from each other. To a first approximation gases do not emit black body radiation. You only see radiation from rotational, vibrational and electronic transitions, and these produce discrete lines. Gas molecules do interact when they collide, so if you increase the number of collisions by making the gas denser and hotter it will start to produce more black body radiation and eventually transition to radiation dominated by the black body spectrum.

Although both the absoprtion and emission spectra are discrete lines, the lines can be quite densely packed. Add a bit of pressure and Doppler broadening and they can blur out to give broad absorption/emission features. However these are still not a black body spectrum.