Pulsar surveys with

1
Pulsar surveys with

• Present surveys, plans for the near future
• Paulo C. Freire, NAIC
• Arecibo, 29th of August 2004

2
In this talk

• Proposed WAPP/ALFA surveys
• General considerations
• Sky coverage
• The ongoing preliminary surveys
• Pointing strategy
• Processing
• Results
• Near-term plans
• Commensality

3
Proposed WAPP/ALFA surveys

• For pulsar surveys, one is interested in using as
  large a bandwidth as possible. For this reason,
  the ALFA pulsar consortium suggested the
  construction of 14 spectrometers capable of
  processing a 300 MHz-wide band, and divide it in
  1024 channels, dumping such spectra every 64 µs.
• The search volume for millisecond pulsars in such
  a survey would be unprecedented.
• This comes at a huge processing/storage cost
  total data produced by the major PALFA survey of
  the Galactic plane 1 PBy.

4
Proposed WAPP/ALFA surveys

• Such capable spectrometers do not exist yet. They
  are being built at Arecibo (B. Sisk and J.
  Hagen), using FPGA technology. The software
  loaded into the gates will make them act as a
  polyphase filterbank.
• Estimated completion Mid-2005
• However, ALFA is already here. And the 4 WAPPs
  now have two correlator boards each. They can
  process a total of 16 100-MHz channels. Can we
  take advantage of it now?
• Yes, we can! This is the aim of the preliminary
  PALFA surveys.

5
Proposed WAPP/ALFA surveys

• The objectives of these surveys are
• Understand observing system, mainly its
  sensitivity, learn how to operate it.
• Develop and end-to-end data processing pipeline,
  including bookkeeping, processing software and
  archival strategies. The latter are very
  important for the main survey that will start
  in about 1-year time.
• If possible, start new science as soon as
  possible.

6
Proposed WAPP/ALFA surveys

• Scientific case for a preliminary shallow survey
• To find brightest undiscovered millisecond
  pulsars (i.e., get the cream of the crop NOW).
  These are the best timers, and they are the ones
  that will benefit most from extended timing
  baselines.
• To find energetic, radio-loud young pulsars in
  the Galactic disk
• Compare yield of this survey with previous
  surveys of similar sensitivity (i.e., Parkes
  multi-beam survey) and with the results of our
  simulations. This data will help optimize the
  large survey.

7
Proposed WAPP/ALFA surveys

• What do we mean by a shallow survey?
• Only 100 MHz bandwidth centered at 1420 MHz,
  divided in 256 channels
• Integrations last only 134 seconds (221 64-µs
  samples), summed polarization mode
• Sensitivity to normal pulsars is slightly better
  than that of the Parkes multi-beam survey
• Much increased sensitivity to millisecond pulsars
  at high DMs

8
Proposed WAPP/ALFA surveys

• Area of the Galaxy to be covered
• Center b lt 0.35, 40 lt l lt 77 (total time
  67 h)
• Anti-center b lt 0.6, 170 lt l lt 186 (total
  time 33h)

9
Proposed WAPP/ALFA surveys

• Things we dont particularly care about
• Band shapes, small gain variations
• Polarization characteristics
• Coma, sidelobes, etc
• Time-constant RFI
• Feed rotation
• Commensality lack of second back-end prevents
  this. So we WAPP it our way
• Things we care about
• Sensitivity

10
Proposed WAPP/ALFA surveys

• Proposed tiling leaves some small holes behind

11
The ongoing ALFA surveys

• The ongoing surveys differ from the proposed
  surveys in two main respects
• Use of sparse tiling pattern (Freire)
• Galactic latitude limit of 1 degree for the
  galactic center survey, instead of 0.35 degrees
• Survey simulations at Cornell (W. Vlemings) and
  McGill (Guiguere) indicate that we can increase
  our detection rate by 50 by making our survey
  sparse, i.e., by fully harnessing the great
  power of Arecibos sidelobes. These are about as
  sensitive as the GBT!

12
The ongoing ALFA surveys

• The proposed tiling pattern would look like this

13
The ongoing ALFA surveys

• What we are really using looks more like this

14
The ongoing ALFA surveys

• The advantages of this are
• We keep the same detection volume for faint
  pulsars,
• We triple (for free!) the detection volume for
  stronger pulsars, because of the 3-times extended
  Galactic latitude range that can be covered at
  that lower sensitivity in the same amount of time
  (more bright MSPs!)
• Dont have to worry about feed rotation at all!
  When we started out survey, the feed rotation was
  still not available, but thanks to the survey
  strategy, this was not a problem.

15
The ongoing ALFA surveys

• Another important innovation in the present
  survey is the real time processing (D. Lorimer)
• We have been using the Arecibo Signal Processor
  (ASP) to FTP the survey observation files to its
  local disks
• Data is then split (each of the 4 WAPPs records
  total power for two beams) and degraded in time
  and frequency by a factor of 16
• Search for slow pulsars in all the DMs then takes
  about 1 second/beam!
• On-line reduction of data speeds up science and
  improves morale!

16
Results

• 1st observing session from the 1st to the 10th of
  August. Second ongoing.
• Everything working perfectly GUI observing mode,
  ALFA IF/LO system, WAPPs. Feed rotation now being
  tested.
• Pulsar MySQL database (J. Hessels) and python
  scripts used to interface with it (Ramachandran,
  Nice, Hessels, Reid, et al.) were installed,
  tested.
• PHP/MySQL web browsing programs (D. Champion)
  made viewing of results of fast processing very
  easy.

17
Results

• Three new pulsars were found to date, after 12
  hours (4 sq. degrees) of observations in the
  center and 12 in the anti-center (8 sq. deg.)
• The first new pulsar is a relatively young 68-ms
  object, not detectable at 430 MHz!

18
Near-term plans

• Understand the observing system better, like
  survey sensitivity (are we finding as many
  pulsars as we should?) and feed rotation (the
  latter will be necessary for dense surveys and
  repeated passes)
• Incrementally improve survey software, implement
  processing software for full resolution data
• Increase data storage and handling capacity and
  computing, in order to prepare full-scale surveys

19
Near-term plans

• Help test 300-MHz back-ends
• Start 300-s, 300-MHz Main survey of the
  Galactic plane most recent simulations indicate
  that there is some slight advantage in doing this
  as compared to two 150-second passes
• First two years make a sparse covering of the
  19-20H Galactic plane visible from Arecibo with
  b lt 5. Simulations indicate that we will be
  able to make 600 of the total expected 900
  detections. Spend following 4 years filling in
  the holes.

20
Commensality

• The proposed tiling scheme for the main survey of
  the Galactic plane should be compatible with
  EALFA requirements
• Some specialized GALFA projects might also make
  good use of it (e.g., recombination line search)
• Pointing scheme should also be fine for SETI
  surveys
• Probably not a good idea for mapping projects
• We are open to discussion on these issues. What
  we are doing now is not necessarily what will be
  done for main survey.

21
It will be challenging to make the 900 detections
we proposed ourselves to do
22
The pulsar consortium thanks

• the people that made it possible
• Jeff Hagen and Bill Sisk for the magnificent work
  with the WAPP and CIMA gui
• Avinash Deshpande, and the electronics department
  for all their great work on the ALFA system
• Arun Venkataraman, for all of his efficient data
  transferring and archiving
• Steve Torchinsky, for managing it all and
  maintaining the irreplaceable ALFA website
• NAIC and Bob Brown for pursuing ALFA, a great
  research instrument, and ATNF for building it.

23
Thank you for your time!

• For questions, comments, etc., contact me at
• pfreire_at_naic.edu, or visit my website
• http//www2.naic.edu/pfreire
• The National Astronomy and Ionosphere Center is
  operated by Cornell University, under a
  cooperative agreement with the National Science
  Foundation.