Decoding FRB energetics and frequency features hidden by observational incompleteness

Fast radio bursts (FRBs) are fierce radio flashes lasting for a few milliseconds from the sky. Although their connection to strongly magnetized neutron stars has been strongly indicated, the exact triggering process and radiation mechanism are still unknown and highly debated. Due to their extremely short duration, the observation of FRBs has long been a difficult task even for large radio telescopes. The difficulty results from the fact that the information obtained in observations is always incomplete, since the telescope always has a limited flux sensitivity and finite operating frequency band. A pressing challenge is to decode the intrinsic features of FRBs from the incomplete observations. Here we establish an efficient methodology to overcome this problem, aiming at effectively correcting for the fluence and frequency cutoffs. Using this method, inverse modeling is performed on a large number of repeating bursts from FRB 20121102A to recover their intrinsic features. It is found that strong bursts intrinsically tend to concentrate their energy in a narrow band, while the spectral range of weak bursts can be either narrow or wide. However, when a weak burst has a broad spectrum, the wing of the spectrum can easily go undetected, resulting in a very narrow spectrum being observed. The narrow spectrum features observed in repeating FRBs are thus an observational selection effect. Underestimation of the burst energy caused by observational cutoffs is also corrected for, and the intrinsic burst energy distribution is re-constructed. It is also found that the bandwidth increases with the increasing central frequency in the Arecibo sample (1.15-1.73 GHz), but such a correlation is not observed in the FAST (1-1.5 GHz) and GBT (4-8 GHz) sample. It indicates the emission pattern of the FRB source might vary across different active periods and frequency bands.