New definition for dual-polarized antenna sensitivity

Paper in proceeding, 2019

Sensitivity is arguably the single most important defining characteristic of a radio telescope. This is because it sets the threshold for what sources can reasonably be detected. Intuitively, sensitivity is the noise level of an antenna referenced to measurements on the sky. So the lower it is, the easier one would expect to make detections of weak sources, and from this it would make sense to call it loosely the 'detection power' of telescopes. But this sensitivity concept does not take into account the polarization of the source or the polarization of the antenna itself, and could thus lead to problems. The traditionally used definition of sensitivity is namely based on single antennas. This antenna based sensitivity is derived assuming that the two polarized antennas that make up radio polarimeters are orthogonal and have identical amplitude gains. In practice these dual-polarized antennas will have different gains and exhibit polarization leakage due to nonorthogonality. To address these more realistic dual-polarized antennas, I introduce a new definition of dual-polarized (or full-polarization) antenna sensitivity. This new definition of sensitivity is a polarimetric generalization of the scalar, system equivalent flux density (SEFD). The new definition generalizes the total sensitivity, and also provides a quantity that represents sensitivity to purely polarized flux. Based on the new SEFD definition, I find that the intuitive notion that identical and orthogonal antennas should typically have better sensitivity than dissimilar, leaky antennas.