Niles said:
I did not know that. Is the probability of the radiation exciting an atom also proportional to the amplitude?
Short answer: ...Ummmm...Yes....
Long answer: Yes. However, this question mixes classical and quantum mechanics. The idea of the wave's amplitude being proportional to the energy density is classical EM. In Quantum Mechanics, the EM wave is propagated in packets of energy/momuntem called photons. The energy of a photon is only dependent upon its frequency. The classical macroscopic EM field can be related to the Quantum Electrodynamic photons using Quantum Electrodynamic theory. However, there are many situations where you can mix classical EM with Quantum Mechanics. The situation you have asked is one situation that we can do so. We can assume that an EM plane wave of amplitude [tex]E_0[/tex] and angular frequency of [tex]\omega[/tex] is incident upon an atom. We can treat the EM wave in the classical sense and the atom in the Quantum Mechanical sense. In this way, we can see that the probability of absorption is directly dependent upon the square of the amplitude, or, equivalently, the energy density of the wave. However, it is also inversely proportional to the square of the difference in energy between the desired transition and the incident wave's photon.
So, the probability to transition to a specific state is proportional to the classical wave's energy density (amplitude squared) and inversely proportional to the square of the difference in energy between the desired excited state (some frequency [tex]\omega_0[/tex]) and the energy of the exciting wave's photon (frequency [tex]\omega[/tex]).
[tex]P_{a \rightarrow b} \propto \left( \frac{E_0}{\hbar \left( \omega_0 - \omega \right)} \right) ^2[/tex]
It is interesting to note that if we want to find the probability of stimulated emission (an EM wave striking an excited atom resulting in the atom emitting a photon of the same frequency as the incident wve) we find that the probability is exactly the same. So the probability for an atom to absorb an EM wave is the same as the probability for stimulated emission.
Anyway, all this is mostly irrelevant to your original question. Classically, the heat from the nonionizing radiation is due to the fact that the EM waves induce currents in conductive medium. These currents give up heat through ohmic loss. Another classical method is the way that microwaves work. Polar molecules, like water, have a slight dipole moment. An EM wave will force these polar molecules to oscillate as they try to line up their dipole moments with the alternating electric field in the EM wave. This kinetic energy gives rise to heat.