top of page
Probing FRB20180916B at High Frequencies with the ATA
Keywords:
#Astronomy #FRB #Radio #Telescope
As part of a summer research internship at the SETI Institute, I joined efforts to study one of the most peculiar repeating Fast Radio Bursts: FRB 20180916B. For some background, FRBS are bright, highly-dispersed, millisecond-duration transients. Unlike typical FRBs, which are often detected once, this source exhibits periodic activity, making it a prime candidate for repeated observation and characterization. My work focused on designing, conducting, and analyzing multi-frequency observations using the Allen Telescope Array, a 20-element radio interferometer.
Project Overview
I conducted more than 28 hours of targeted observations of FRB 20180916B, in addition to experimental runs on other FRBs and known pulsars for calibration. I developed Python scripts to automate a multi-step process: tuning the array to specific frequency bands, slewing the antennas to the source after calibrating on a bright pulsar, and managing data collection in manageable 30-minute segments due to storage constraints.
Observations targeted two distinct frequency bands centered at 1236 MHz and 8000 MHz, covering both well-studied and underexplored parts of the spectrum. Following each session, I processed the data using the SPANDAK pulse detection pipeline. This involved validating the system with pulsar detections and then visually inspecting hundreds of dynamic spectrum plots to identify potential FRB signals.
Why Null Results Still Matter
Despite rigorous observation protocols and a detection threshold requiring a signal-to-noise ratio greater than 10, no bursts were detected during the campaign. While this may seem disappointing, the absence of detections provides valuable constraints on the emission properties and activity windows of FRB 20180916B. More importantly, null results are still scientifically significant, they refine models, inform future observation windows, and motivate broader frequency coverage. This project gave me first-hand experience in radio instrumentation, signal processing, and the patience required in exploratory astrophysics.
And we are not completely without results! As our investigation still provides insight into the luminosity and energy of our dataset. We begin this by calculating the System Equivalent Flux Density (SEFD) for each observation, utilizing the flux models from Perley & Butler (2017) to provide predicted flux densities at the central frequencies of each observation. This process involved calculating the average SEFD for all 20 antennas with 2 polarizations within the beamformer and subsequently dividing it by the number of elements in the array (in this case, 20), thereby determining the SEFD for each observation with the following equation:
We find that the average SEFD for the beamformer at 1236 MHz is 534.423 Jy and at 8000 MHz it is 1208.825 Jy. Further, we use these values to get to the weakest signal in Jy our experiment would detect through the radiometer equation, with a signal-to-noise ratio of 10, an average FRB duration of 𝜏 = 0.001 seconds, and bandwidth of Δ𝜈 = 672 MHz(assuming FRB burst bandwidth is limited by capabilities of the ATA):
The resulting minimum detectable flux density in the 1236 MHz tuning is 4.62 Jy and at the 8000 MHz tuning is 10.427 Jy. This information helps us assess whether the lack of detected pulses is due to the instrument’s limitations or the actual absence of signals from the source. In our case, and considering that most bright radio sources are usually between 1-100 Jy, the obtained values align with expectations.
For a more academic and detailed approach of my work, you can read my end of summer paper:
bottom of page