Missing the Forest for the Trees

1. Missing the Forest for the Trees Science with the Byrd Green Bank Telescope K.K. Dyer - NRAO

2. The zero-spacing problem: • Interferometers work by adding signals between antennas. • Short baselines measure large scale flux. • Long baselines provide high resolution -- they sample small scale structure. • Therefore there is an angular size scale beyond which you cannot observe structure.

5. Observational Status Summary ....structures on angular scales significantly larger than the fringe spacing formed by the shortest baseline are not measured. No subsequent processing can fully recover this missing information, which can only be obtained by observing in a smaller array configuration, using the mosaicing method, or with an instrument (such as a large single antenna) which provides this information

6. How much do you lose? • Smallest scale information = /D • D = distance between antennas • For the VLA in D array at 20 cm ~ 17 arcminutes

7. What kind of observations does this affect? • Nearby galaxies • Galaxy clusters: Coma, Virgo • Supernova remnants • Planetary nebulae • Galactic plane • HI

8. CARMA and ALMA • 12 meters, 90 GHz = 1 arcminute • 15.4 meters, 490 GHz = 10 arcsecond So this will be a problem EVERYWHERE

9. Smallest scale measured is different at each frequency! • 6 cm: 2.7 Jy • 20 cm: 18.5 Jy What we should get: 7 Jy 15 Jy

10. Single dish measurements

11. What is in a name? • G327.6+14.6 • RA 15 02 50, DEC -41 56 • SN1006 X-ray image, Jack Hughes

12. Technical Details How low can you go? Dec -41º • Choose a method

13. Using GBT data to inform the VLA data • Fake synthesis • Joint deconvolution • Merge (add in the image plane) • Single dish as starting image GBT 20 cm observation of SN1006

16. Listening to your elders... total collecting area antenna efficiency system temperature bandwidth VLA=GBT small dishes vs. large dishes Time to reach the same point source sensitivity at 1.4 GHz 6 years Proof by Contradiction

17. final image

18. constant noise fallacy Even at the VLA we don’t observe to constant noise: Each configuration gets x 4 20 seconds per GBT beam = 8 hours GBT time A config 20 hours B config 6 hours C config 2 hour D config 1 hour

19. goal: constant signal to noise Don’t observe longer once you’ve hit the systematics. 0.5 seconds per beam= 2 hours of observing time Note: Fourier plane method make everything four times worse! We achieve 50:1 SNR in 1 sec

20. The result: VLA contributes 300 mJy/beam GBT contributes 10 mJy/beam on 8.5’ scales Noise is different on different scales!

21. final image

23. cross section 2

24. cross section 1

25. final image

26. low level emission

27. GBT+ VLA BCD 20 cm, Cornwell, Holdaway, Dyer

28. On to bigger fish:M82

29. Spectral energy distribution of M82

31. Conclusions: • Interferometers do high resolution well, total flux poorly • Worry about small spacing scale • SN1006 -- optical & x-ray • On to M82 and FarIR-radio!

32. Science issues affected by zero spacing problem Supernova remnants: • Testing particle acceleration • Determining magnetic field

33. Example 1:Cosmic rays, Particle acceleration

34. Example 2:The magnetic field across the shock in SN1006 • Critical example of shock acceleration • Disagreement over nonthermal x-ray emission • Confirm the magnetic field GBT 20 cm observation of SN1006