A radio technosignature search towards Proxima Centauri resulting in a signal of interest

Proxima Centauri b’s discovery has intensified interest in the nearest potentially habitable exoplanetary system. Although M-dwarf hosts present challenging environments (strong stellar activity, frequent flares and coronal mass ejections), modeling suggests Proxima Centauri b could sustain an atmospheric water cycle and possibly maintain habitability under certain conditions. The system is also central to Breakthrough Starshot concepts, which envision laser-driven probes returning data from Alpha Centauri within decades. Despite proximity, few technosignature searches have targeted Proxima Centauri historically, leaving a gap this work addresses. The purpose here is to conduct a sensitive, wide-band radio search for engineered narrowband emissions consistent with extraterrestrial technologies, quantify detection limits, and rigorously filter terrestrial interference.

Previous SETI efforts in the southern hemisphere (e.g., Project Phoenix searches of 202 solar-like stars and surveys of 176 bright stars in the 1990s) did not include the faint M-dwarf Proxima Centauri. Until recently, there had been no optical technosignature searches for this target; a search of archival HARPS spectra (2004–2019) constrained laser emission from Proxima Centauri to less than ~120 kW without detection. Breakthrough Listen has undertaken comprehensive SETI campaigns since 2016, including prior Parkes 10 cm (2.60–3.45 GHz) observations and Green Bank Telescope surveys, establishing methodologies and sensitivity baselines that this study improves upon.

Instrumentation and setup: Observations used the CSIRO Parkes (Murriyang) 64-m telescope with the Ultra-Wideband Low (UWL) receiver covering 0.704–4.032 GHz (~3.3 GHz bandwidth). Typical system temperature was ~22 K and SEFD ~28 Jy over ~60% of the band. Data were recorded via the Breakthrough Listen Digital Recorder at Parkes (BLDR/BLDRP), producing high spectral resolution products (channel resolution ~3.81 Hz; time integrations ~16.8 s) stored in filterbank format; total data volume for 64 observations was ~117.81 TB. Observation campaign and strategy: Proxima Centauri was observed from 2019-04-29 to 2019-05-04 for a total of approximately 26 hours. An on–off pointing strategy (to discriminate RFI) was employed, with on-source segments of ~30 minutes and off-source segments of ~5 minutes, and cadences extended beyond the typical six-pointing pattern to 12–17 pointings. This strategy increases sensitivity to persistent or long-duration signals (≥30 minutes) and enables improved RFI rejection across multiple on/off cycles. Signal search pipeline: The turboSETI Doppler search code scanned the 0.7–4.0 GHz band for narrowband signals exhibiting Doppler drift due to relative accelerations. Hits (detections above a 5σ threshold) were collated and passed to an event-level filter requiring presence in on-source and absence in off-source pointings. Further filtering included visual inspection of dynamic spectra, verification across cadences, and frequency cross-checks against known local transmitters (cellular, satellite, broadband internet) and regulatory allocations (Australian Radiocommunications Spectrum Plan; ACMA licensing database). Data conditioning handled bandpass instabilities and sinusoidal baseline structures. Sensitivity and detectability: The minimum detectable flux F_min for narrowband signals was computed from telescope parameters and integration time; from this, the minimum detectable equivalent isotropic radiated power (EIRP_min) at d = 1.301 pc (Proxima Centauri) was derived. Using T_sys = 22 K and observation settings, F_min ≈ 9.2 yJ Hz and EIRP_min ≈ 1.9 GW, a factor ~3.6 better than earlier Parkes 10 cm observations and ~1.6 better than comparable GBT L/S-band searches of a nearby solar neighbor. Expected Doppler drift ranges were estimated from rotational and orbital dynamics of Earth and the Proxima system to set search bounds and avoid missed detections. Event identification and vetting: Events were defined as narrowband detections present in on-source but absent in off-source scans. Candidate events surviving automated filters underwent additional multi-pointing verification (extending to 6-pointing cadences where possible), visual scrutiny, and exclusion if falling within registered transmitter bands or otherwise consistent with terrestrial/satellite RFI.

- The pipeline detected millions of narrowband hits across the band; a frequency histogram showed 57% of hits within ranges occupied by registered terrestrial/satellite transmitters near Parkes. - Drift-rate distribution among hits: ~10% positive, ~15% negative, and ~75% near zero drift, with stronger signals generally associated with nearby ground-based transmitters appearing in both on- and off-source pointings. - Total events (detections in on-source but absent in off-source) numbered 5,160 after initial filtering. - Only one event, designated BLC1 (Breakthrough Listen Candidate 1), passed all automated and visual filters: detected at 982.002571 MHz with an observed Doppler drift rate of approximately −0.238 Hz s⁻¹, persisting over more than 2.5 hours and present only during pointings towards Proxima Centauri. It lay outside known local RFI allocations for the site, despite the broader band including aeronautical radionavigation allocations in that region; no registered transmitters operated at that exact frequency within 1,000 km. - Sensitivity: The search achieved F_min ≈ 9.2 yJ Hz and EIRP_min ≈ 1.9 GW at 1.301 pc, improving limits relative to prior Parkes and GBT surveys, thus probing lower-power putative transmitters than previously reported for this target. - Although BLC1 exhibited properties consistent with a putative technosignature, a companion analysis ultimately attributes it to unusual locally generated interference.

This study demonstrates a highly sensitive, wide-band technosignature search towards the nearest stellar neighbor, with rigorous RFI discrimination via extended on–off cadences and regulatory cross-checks. The prevalence of zero-drift and strong-hit signals associated with registered transmitters underscores the importance of multi-pointing strategies and thorough RFI databases. BLC1’s survival through stringent filters highlights that rare, subtle interference can mimic hypothesized extraterrestrial signals; comprehensive follow-up is critical to avoid false positives. The achieved EIRP_min limit of ~1.9 GW places tighter constraints on continuous or long-duration narrowband transmitters in the Proxima Centauri system over 0.704–4.032 GHz than prior work, informing future survey design and target prioritization. Continued multi-instrument, multi-wavelength campaigns will be necessary to further reduce parameter-space gaps and to replicate or refute signals of interest across observatories and epochs.

The work presents one of the most sensitive radio technosignature searches of Proxima Centauri to date, identifying a single compelling signal of interest (BLC1) that, upon deeper analysis in a companion paper, is attributed to local interference. The survey improves detection limits (EIRP_min ~1.9 GW) relative to previous efforts, refines event-vetting procedures, and quantifies the distribution of hits and events across a wide band. Future efforts should: (i) expand spectral coverage to frequencies beyond 0.704–4.032 GHz and to other wavelengths (optical, IR, X-ray), (ii) employ coordinated, simultaneous multi-facility observations for rapid verification and RFI discrimination, (iii) explore sensitivity to shorter-duration and higher-drift-rate signals, and (iv) continue improving RFI characterization and databases.

- Spectral coverage is limited to 0.704–4.032 GHz; no contemporaneous optical/IR/X-ray technosignature searches were performed. - Observation strategy prioritizing ~30-minute on-source integrations reduces sensitivity to transient signals of duration less than ~30 minutes. - Despite robust filtering, terrestrial/near-Earth RFI remains a confounding factor; rare interference can pass standard filters, as illustrated by BLC1. - Single-target, short-duration campaign limits generalizability; wider samples and repeated epochs are needed to constrain occurrence rates. - Drift-rate search bounds focused on expected astrophysical dynamics; extremely high drift rates from hypothetical fast-orbiting spacecraft or atypical transmitters were not comprehensively explored.