A search for Extraterrestrial Intelligence (SETI) toward the Galactic Anticenter with the Murchison Widefield Array
AA search for Extraterrestrial Intelligence (SETI) toward theGalactic Anticenter with the Murchison Widefield Array
S.J. Tingay , C.D. Tremblay , & S. Croft ABSTRACT
Following from the results of the first systematic modern low frequency Search for Extrater-restrial Intelligence (SETI) using the Murchison Widefield Array (MWA), which was directedtoward a Galactic Center field, we report a second survey toward a Galactic Anticenter field.Using the MWA in the frequency range of 99 to 122 MHz over a three hour period, a 625 sq. deg.field centered on Orion KL (in the general direction of the Galactic Anticenter) was observed witha frequency resolution of 10 kHz. Within this field, 22 exoplanets are known. At the positions ofthese exoplanets, we searched for narrow band signals consistent with radio transmissions fromintelligent civilisations. No such signals were found with a 5 σ detection threshold. Our sample issignificantly different to the 45 exoplanets previously studied with the MWA toward the GalacticCenter (Tingay et al. 2016), since the Galactic Center sample is dominated by exoplanets de-tected using microlensing, hence at much larger distances compared to the exoplants toward theAnticenter, found via radial velocity and transit detection methods. Our average effective sensi-tivity to extraterrestrial transmiter power is therefore much improved for the Anticenter sample.Added to this, our data processing techniques have improved, reducing our observational errors,leading to our best detection limit being reduced by approximately a factor of four compared toour previously published results. Subject headings: planets and satellites: detection – radio lines: planetary systems – instrumentation:interferometers – techniques: spectroscopic
1. INTRODUCTION
This paper continues efforts to utilise theMurchison Widefield Array (MWA: Tingay etal. 2013) for the Search for Extraterrestrial In-telligence (SETI). Previously, we utilised MWAobservations primarily targetted at the detectionof astrophysical spectral lines along the GalacticPlane, toward the Galactic Centre, over a 400 sq.deg. field, to undertake an opportunistic search fornarrow band SETI signals in the frequency range103 to 133 MHz (Tingay et al. 2016). The readercan refer to Tingay et al. (2016) for the generalcontext of SETI research at radio wavelengths,as well as some brief history specifically for low International Centre for Radio Astronomy Research(ICRAR), Curtin University, Bentley, WA 6102, Australia Department of Astronomy, University of CaliforniaBerkeley, Berkeley, CA, USA frequency radio observations. We will not repeatthis context here. An up-to-date description ofSETI surveys conducted at cm-wavelengths canbe found in Enriquez et al. (2017) and an excellentgraphical summary of previous radio SETI exper-iments can be found in Gray & Mooley (2017)(their Figure 9).To briefly reiterate the unique capabilities ofthe MWA described in Tingay et al. (2016), wecover a frequency range for SETI that has beenpoorly explored to date (in particular from a pris-tine radio quiet environment) and the MWA hasan extremely wide field of view (up to 1000 sq.deg., depending on frequency). This means thatvery large numbers of SETI targets can be ex-amined in any given observation in a unique fre-quency range. Gray & Mooley (2017) place theTingay et al. (2016) results in the overall con-text of SETI experiments and show that they are1 a r X i v : . [ phy s i c s . pop - ph ] F e b ighly competitive in this frequency range.We did not detect any SETI signals from the 45exoplanets known in our previous Galactic Cen-tre field, with a best limit on estimated Equiva-lent Isotropic Radiated Power (EIRP) of approx-imately 4 × W from GJ 6676c at a distanceof 6.8 pc. Only 4/45 exoplanets in our GalacticCentre sample are closer than 50 pc. The remain-ing 41 exoplanets are all more distant than 1 kpc,since the sample is dominated by exoplanets de-tected using microlensing techniques. Therefore,the vast majority of EIRP limits in the GalacticCenter field sample are more than four orders ofmagnitude higher than our best limit in that field.In this paper, we examine a different field withthe MWA, generally toward the Galactic Anticen-ter (centred on Orion KL), where the known ex-oplanets are mainly detected using radial velocityor transit techniques. The 22 known exoplanets inthis field consist of 15/22 closer than 100 pc with12/22 closer than 50 pc. The closest exoplanets inthis field are BD-06 1339b/c, at a distance of 5.32pc, 20% closer than the closest exoplanet in ourGalactic Center field. Thus, on average, our sensi-tivity to extraterrestrial transmitted radio poweris much higher for our observations in the GalacticAnticenter field, compared to the Galactic Centerfield.The MWA is participating in a new wave ofSETI experiments of various types, summarised(MWA included) in Worden et al. (2017) andlargely driven by the new Breakthrough Listenprogram. Breakthrough Listen is gaining momem-tum and has published the first results from a sur-vey of 692 nearby stars between 1.1 and 1.9 GHzwith the Green Bank Telescope (GBT) (Enriquezet al. 2017). The upper limits on EIRP from theseearly BL results are of the order 10 W, compara-ble to the limits we present in Tingay et al. (2016)from the MWA. Thus, interestingly, the best GBTand MWA limits are very similar but cover fre-quencies that are an order of magnitude different(100 - 200 MHz vs 1 - 2 GHz). There is no overlapbetween the 692 targets of Enriquez et al. (2017)and the sample of 22 exoplanets reported here.Also, the GBT survey searched for signals witha bandwidth of a few Hz, whereas our survey hasa resolution of 10 kHz. The signal-to-noise ratioof a narrow-band signal (if corrected for Dopplerdrift) will be much lower in a 10 kHz band than it will be in one with resolution that approximatelymatches the bandwidth of the signal. However, al-though transmitter power requirements might in-fluence an extraterrestrial civilization to prefer tobroadcast narrow-band signals, it is by no meanscertain that they would choose to do so. For ex-ample, spread spectrum (wide band) communica-tions techniques have been considered in the pastby some authors (Messerschmitt 2012). In addi-tion, the coarser frequency resolution of our surveymeans that a typical signal will stay confined to asingle channel during the integration time of ourobservations, meaning that we can simply look foroutliers in individual channels, rather than per-forming a full Doppler drift search. Experimentswith the MWA are therefore complementary tothose with the GBT or other facilities at muchhigher frequency resolutions.In the following sections we describe our obser-vations and data analysis, and provide discussionof our conclusions.
2. OBSERVATIONS AND DATA ANAL-YSIS
Observations with the MWA took place on 21November 2015, centered on the Orion Kleinmann-Low Nebula (Orion KL), as described in Table1. Dual-polarisation data were obtained in a30.72 MHz contiguous bandwidth. A two stagepolyphase filterbank splits the bandwidth into24 × ×
10 kHz“fine” spectral channels. Observations were takenin two minute segments over a total of 180 min-utes, a field of view (primary beam) FWHM of ∼ ◦ at a resolution (synthesized beam) of 3.2 (cid:48) .However, only 625 deg were imaged and searched,within the most sensitive region of the primarybeam.The data were imaged and calibrated as part ofthe molecular line survey of the Orion MolecularCloud complex (Tremblay et al, in prep) using thepipeline described in Tremblay et al. (2017). Byusing WSClean (Offringa et al. 2014) we imagedeach coarse and fine spectral channel using Briggsweighting “-1” to compromise between image res-olution and sensitivity. Images from each coarsechannel were used to determine the effects of theionosphere on the source positions within the field2f view, and a single linear correction was basedon a comparison with GaLactic and Extragalac-tic All-sky Murchison Widefield Array (GLEAM)survey point source catalog (Hurley-Walker et al.2017) stacked images at 103 MHz. After correc-tion, the residual astrometric uncertainty is 1 (cid:48)(cid:48) inright ascension and 5 (cid:48)(cid:48) in declination; both ofwhich are significantly smaller than our beam.The edges of each coarse channel suffer fromaliasing, requiring a number of fine (10 kHz) chan-nels on each coarse channel edge to be flagged.This resulted in 2400 (78%) of the 3072 fine spec-tral channels being imaged. The central fine chan-nel of each coarse channel was flagged, as theycontain the DC component of the filterbank. Au-tomated flagging of radio frequency interference(RFI) was performed using AO Flagger (Offringaet al. 2012) in each (2-minute) snap-shot obser-vation prior to the visibilities being imaged andstacked. As noted in Tingay et al. (2016) and Tin-gay et al. (2015), the application of AO Flagger de-tects RFI onn single baselines and over short inte-gration periods, identifying and removing RFI cor-responding to thouands of Jy, three to four ordersof magnitude higher than the levels being probedin this paper and corresponding to known terres-trial transmitter frequencies (generally in the FMband).A search of the field for exoplanets, based on theKepler catalogue (Akeson et al. 2013), returned17 known planetary systems containing 22 exo-planets. The systems are listed in Table 2 andare shown in relation to the MWA field of view inFigure 1.The MWA data cube was searched at the loca-tions of each of these exoplanet systems and spec-tra were extracted from these locations. No sig-nificant narrow band signals were detected abovea 5 σ level in any of these spectra. Table 2 lists theRMS from the spectra at each of the exoplanet sys-tem locations and the corresponding 1 σ limits onthe inferred isotropic transmitter power requiredat the distance of the exoplanet system. Figure2 shows an example spectrum extracted from theMWA data cube, representing the data for exo-planets BD-06 1339b/c.
3. DISCUSSION
GJ 3323b and c, at 5.32 pc, are the closest ex-oplanets in Table 2, 20% closer than the closestexoplanet we observed in Tingay et al. (2016), GJ667c. The measured RMS toward GJ 3323b/c isapproximately half of the RMS we measured to-ward GJ 667c. Thus, our limit on EIRP for GJ3323b/c is approximately four times lower than weprevsiouly estimated for GJ 667c, 10 W.In terms of the sample in this paper, the me-dian distance ( ∼
50 pc) is a factor of approximately40 lower than the median distance of the systemsin Tingay et al. (2016) ( ∼ ∼ W is still high compared to the highestpower low fequency transmitters on Earth, eventhough this EIRP limit is plausibly within reachof Earth-based transmitters at frequencies a factorof ten higher (Enriquez et al. 2017).We note that the primary goals of the observa-tions utilised here are searches for radio recombi-nation lines and molecular lines. Tremblay et al.2018a (in preparation) and Tremblay et al. 2018b(in preparation) report both types of astrophysicallines, including three instances of lines that haveno identified molecular transition but are plausi-bly molecular lines rather than SETI signals, giventheir coincidence with evolved stars.We note that we only consider explicitly herethe positions of known exoplanets. Across theTable 1: MWA Observing Parameters
Parameter ValueCentral frequency 114.56 MHzTotal bandwidth 30.72 MHzNumber of imaged channels 2400Channel separation 10 kHzSynthesized beam FWHM 3.2 (cid:48)
Primary beam FWHM 30 degreesPhase center of map (J2000) 05h35m, –05d27mTime on source 3 hoursDate Observed 21 November 2015
The MWA is continuing to build toward larger-scale SETI experiments, in collaboration with theBreakthrough Listen team.In addition to searches of wide-field image cubessuch as the one reported here, the MWA has thecapability to beamform using voltages on specifictargets of interest. Voltages provided by the Volt-age Capture System (VCS; Tremblay et al. 2015)can be processed into data products including in-coherent and coherent beams. The former enablesa search of a single pixel corresponding to the pri-mary beam of an individual MWA tile; the latterprovides (in the case of the MWA) an order of magnitude improvement in sensitivity, for a beamcomparable in size to the synthesized beam pro-duced by the correlator. For frequency resolutionsimilar ( ∼
10 kHz) to that of the search reportedhere, forming an image cube using the correlatoris tractable in terms of computational expense anddata volume, but as frequency resolution improves(and the number of channels increases), beam-forming in the direction of targets of interest be-comes preferable.The VCS has been used for fast transientsearches and pulsar studies (e.g. Bhat et al.2016; McSweeney et al. 2017), among other ap-plications. The two main current limitations tothe VCS are the network infrastructure at theMurchison Radio-astronomy Observatory (MRO),limiting the use of the system to just a few hoursper week, and the manner in which the data arechannelized (limiting the frequency resolution thatcan be obtained). A new 100 Gbit/s link nowoperational from MRO to Curtin University over-comes these limitations, and the installation ofa new Breakthrough Listen computational facil-ity at Curtin University similar to those deployedat GBT and Parkes (MacMahon et al. 2017) willenable a high frequency resolution, real-time com-4
System RA Dec Dist. (pc) RMS (Jy/beam) P (10 W)BD-06 1339b 05h53m00.28s − < − < < − < − < − < − < − < − < − < < < < < < − < < − < − < . × WASP-35b 05h04m19.63s − −− WASP-49b 06h04m21.46s − < . × WASP-82b 04h50m38.56s +01d53m38.1s 200 0.46 < . × Table 2: The 22 known exoplanets in the MWA field of view. Column 1 - Exoplanet system; Column 2 -right ascension (deg); Column 3 - declination (deg); Column 4 - distance (pc); Column 5 - RMS (Jy); andColumn 6 - upper limit on isotropic transmitter power in units of 10 W. The distance of WASP-35b is notknown.mensal SETI search to be performed. This newinstrument will not only provide improved SETIsearch capabilities on MWA, but will enhance theoverall capabilities of the telescope for fast tran-sient, pulsar science, molecular line, and solarstudies, and is expected to be installed and oper-ational in 2018.The MWA’s wide field of view and broad rangeof science cases means that a commensal user canbuild up a deep all-sky survey without ever point-ing the telescope. Over the course of a year, almostthe entire visible sky is covered to a depth of atleast several hours, with some smaller deep fieldscovered to an order of magnitude or more greaterdepth. By accessing the data before fine chan-nelization is performed, a substantial improve-ment over the current 10 kHz frequency resolu-tion can be obtained, enabling a much more sen-sitive search for narrow band ( ∼ Hz bandwidth)transmitters, as well as improved discriminationbetween natural spectral lines, RFI, and artificialsignals of interest.
4. Acknowledgements
We thank the anonymous referee for provid-ing comments that significantly improved themanuscript. The authors would like to acknowl-edge the contribution of an Australian Govern-ment Research Training Program Scholarship insupporting this research. This work was sup-ported by resources provided by the Pawsey Su-percomputing Centre with funding from the Aus-tralian Government and the Government of West-ern Australia. This scientific work makes use ofthe Murchison Radio-astronomy Observatory, op-erated by CSIRO. We acknowledge the WajarriYamatji people as the traditional owners of theObservatory site. Support for the operation of theMWA is provided by the Australian Government(NCRIS), under a contract to Curtin Universityadministered by Astronomy Australia Limited.We gratefully acknowledge the support of NASAand contributors of SkyView surveys. Funding forBreakthrough Listen research is provided by the5ig. 2.— MWA spectrum for BD-06 1339b/c. Areas of flagged fine channels on coarse channel edges areevident as gaps in the spectrum.Breakthrough Prize Foundation . Facility:
MWA.
REFERENCES
Akeson, R.L. et al. 2013, PASP, 125, 989Bhat, N.D.R., Ord, S.M., Tremblay, S.E., Mc-Sweeney, S.J. & Tingay, S.J. 2016, ApJ, 818,86Enriquez, E. et al. 2017, ApJ, 849, 104Gray, R.H. & Mooley, K. 2017, AJ, 153, 110Hurley-Walker, N et. al., 2017, MNRAS, 464, 1146Messerschmitt, D.G. 2012, Atca Astronautica, 81,227MacMahon, D.H.E. et al. 2017, arXiv:1707.06024 http://breakthroughprize.org McSweeney, S.J., Bhat, N.D.R., Tremblay, S.E.,Deshpande, A.A. & Ord, S.M. 2017, ApJ, 836,224Offringa, R.A. et al. 2014, MNRAS, 444, 606Offering, R.A. et al. 2012, MNRAS, 422, 563Tingay, S.J., Tremblay, C., Walsh, A. & Urquhart,R. 2016, ApJ, 827, L22Tingay, S.J. et al. 2015, AJ, 150, 199Tingay, S. J. et al., 2013, PASA 30, 7Tremblay, C. et al. 2017, MNRAS, 471, 4144Tremblay, S. et al. 2015, PASA, 32, 5Worden, P. et al. 2017, Acta Astronautica, 139, 98