I believe the answer is C.
In AAS, the analyte emission spectra (from hollow cathode lamp) is shown at the analyte, which absorbs the radiation. Thus we are measuring the absorption of the emission spectra. This emission spectra is not emitted by the analyte in sample, but it is emitted by standardized analyte in the cathode lamp.
One thing to stress is that AAS is know for discrete emission lines. I think it is incomplete to say it is absorption of visible and UV light, which can mean a continuous spectrum is being generated at the source.
In AES, we are not measuring absorption, we are measuring the amount of emission.
I think that disagrees with the definition of analyte though. Analyte is specifically the thing you are measuring - in this case the concentration of. I agree that neither B nor A is a perfect answer either, because, as you say, you are measuring discrete wavelengths, probably only one, in which case UV and visible light are not both used, so UV or visible light is probably a better way of saying it. I still think to talk about the "emissions of the analyte" specifically means what is sitting in the atomiser, not what is in the HCL. That's my opinion.
I am not convinced by that argument, but I will agree that the options are ambiguously worded. If we take the interpretation that analyte refers to atoms in the atomiser, option C wouldn't make any physical sense. Absorption of emission would imply a decrement in emission by analyte in the atomiser, but such thing cannot be measured since initial emission intensity cannot be measured, thus absorbance cannot be calculated. In my opinion, it is a poor use of the word 'analyte'. If we assume used the word 'analyte' as a loose term (i.e. 'analyte in the instrument' = HCL, where analyte = metal of interest), it is exactly the principle of AAS. Another thing to consider is that the analyte depends on which HCL you use, so their loose use of 'analyte' may be justifiable.
I'm just a bit caught up on metal atoms absorbing visible and UV light. I still believe the most important thing we learn about AAS is that discrete spectra are emitted and absorbed, giving rise to its selectivity and sensitivity. The emission range is less important, and a question should not be testing this aspect.
@Water, do you have the solutions to this problem?