BINGO — Baryon acoustic oscillations In Neutral Gas Observations

BINGO project  is about building a  radio telescope in such a way that systematics are controlled from the beginning, permitting the extraction of the HI signal via a well controlled calibration, and at a controlled cost.

BINGO will observe along 2 years (1 year on source) to estimate the acoustic scale at a 2% uncertainty level, which will permit to estimate the equation of state of dark energy w at 16% uncertainty level.  

The main goal is to survey  a region of  200 deg²  of the sky with an RMS of  0.1 mJy and spectral stability.  Dark energy is best probed at a later stage of the Universe, and BINGO’s motivation to measure  HI fluctuations at a redshift of z < 0.5.   As shown above, the aimed signal is expected to be of the order of 0.1 mK, foreground contamination is expected to be of the order of 5 K. The contribution of the detecting system to the signal is foreseen to be T_sys ~ 50 K, which won’t be a problem if stable, and all efforts have been done to find a clean site in terms of RFI.

the telescope: BINGO will be a single dish telescope, static, taking advantage of the Earth rotation to map a 10⁰ strip of the sky,  using a focal plane horn array.  More specifically,  will be composed by a two-mirror crossed-Dragone design, an under illuminated primary dish of 40 m, a secondary dish slightly smaller and 50 horns of 4 meters length. The needed extended field of view argues for a large number of horns, and in order to accommodate such number of horns, a large focal distance is needed, which is foreseen to be f/D=3.

frequency band: BINGO will operate between 960 MHz -1260 MHz, allowing to detect  21 cm emission at a redshift from 0.12 to  0.48, a band that also should be clean in terms of  RFI from communications.

spectral and angular resolution:  BINGO will have a spectral resolution of 1 MHz and an angular resolution of 40 arcmin.

receivers: will be standard correlation receivers, to reduce 1/f noise, the output being the difference between the signal from a science beam and a signal from  a known source. To control cost, ambient temperature Low Noise Amplifiers (LNA) will be used,  the total system temperature foreseen to be  T_sys ~ 50K. Much of the receiver will be in digital hardware. Gain stability needs to be guaranteed for periods of 20 minutes —  corresponding to the time of a  transit in the sky.

sensitivity: the receivers need to be extremely stable, at most a noise of 0.1 mK in each 1 MHz frequency channel.  In order to be able to detect the aimed fluctuations of  0.1mK, it is necessary to integrate for years, considering that BINGO is a transient instrument, each region of the sky being observed 20 minutes per pass.

Projected BINGO results

Below is presented the simulated HI signal for BINGO, but in this case assuming a telescope beam of 60 arcmin (will be 40 arcmin) and a frequency resolution of 1 MHz. The aimed fluctuations will be pinpointed as the difference between the signal and an average represented by the dashed line.

The foreseen 5 σ  detection will be achievable after 1 year of integration on source, and will allow a measurement of the acoustic scale with  2.4% uncertainty, and constraint the dark energy equation of state with 16%  uncertainty.

The HI angular power spectrum at z = 0.28 – corresponding to the centre frequency of BINGO – along with projected errors on the reconstruction from the proposed BINGO survey with 50 feedhorns and 1 year of on-source integration time covering 2000 deg² . Included as an inset is the power spectrum divided by a“smooth spectrum” to isolate the BAOs —  Figure and caption from Baytte et al 2016.

Acoustic scale measurements allow to further constrain BAO Hubble diagram, ie, angle-distance  redshift relation, and the projected angle-averaged distance from BINGO measurements is plotted along with results from other works, see below

Adapted from Baytte et al 2013

For that it is mandatory a good calibration in terms of foregrounds and systematics,  below are some of  the simulations on both aspects, assuming a telescope as described before. The simulations are from Bigot-Sazy et al (2015), details therein:

Above a simulation showing  galactic and extragalactic  foregrounds at 1GHz, the sky region that will be probed by BINGO selected in white solid lines.  Below,  a simulation showing the foreseen signal for a drift scan strip (corresponding to the probed region, grey represents non observable regions) HI emission in the top plot, galactic and extragalactic foregrounds in the bottom plot.

Above, a simulation showing, for the same region, and for a drift scan, the signal from instrumental noise, the graphic of the top representing  the thermal noise, and that of the bottom the 1/f noise.  Worth noting the differences in intensity scales among the different simulations.

The astronomical and instrumental foregrounds unwanted signals may be cleaned using different methods, below it is shown the results via parametric fitting and principal component analysis (PCA). The simulations refer to instrumental noise (thermal noise and 1/f noise) and foregrounds (galactic synchrotron emission plus point extragalactic sources) From top to the bottom, the strips refer to:

  • 1st strip —  HI signal
  • 2nd 3rd 4th strip — recovered HI signal after one, three and seven principal modes were removed
  • 5th strip — recovered HI signal after parametric fitting applied

These simulations show that principal analysis component can retrieve reasonably well the HI signal from a contaminated map, in this case after 7 principal modes removed.

Besides BAO

Besides the HI intensity mapping,  BINGO will be a fantastic telescope for transient science. One of the astronomical subjects that may take advantage of such instrument  are the so called Fast Radio Bursts (FRB).

First Fast Radio Burst detection, Lorimer et al 2007

Very little is known about FRB, its origin to be determined. They are bright and very short (millisecond scale) phenomena, among the few registered events (20 up now) only one repeats in a non regular basis. FRB signal is very broad in frequency, the dispersion measure being much higher than what is expected if related with galactic material, favouring an extragalactic origin.

Gemini Observatory/AURA/NSF/NRC

Interesting enough, very recently the only repeating source has been associated with a galaxy (Keane et al 2016) . But apart from this particular object, the origin and location of these objects are a mystery.

BINGO being a transient telescope, will be in position to make a difference in FRB knowledge.  It is estimated that it could be possible to detect FRBs at a rate of 1 per week, with additional electronics being added to achieve high time resolution — not necessary for BAO detection. In a future  phase  small outrigger optical telescopes can be added in order to measure real time accurate positions and posterior optical identification.