I have been trying to calculate the rotational constants (B) for CO and CO2 from IR Vibrational-Rotational spectra.

I know that for CO the peak spacing is approximately 2B (ignoring effects due to centrifugal distortion) however for CO2 it seems the peak spacing represents 4B instead when I compare my results to literature?

Does anybody have any explanation as to why this is, or could provide reading on this subject to explain why the peak spacing for CO2 is 4B? Thanks!