LOFAR Long Baseline Calibration Commissioning anderson@mpifr-bonn.mpg.de On behalf of LOFAR and the LLBWG 1/31
No, No Fringes On Long Baseline Yet... I hate pretending to be an optimist when writing abstract proposals Effelsberg Exloo testing since 2008 Mar in low band Single baseline sensitivity about 50 Jy in 10 s Groningen correlator replaced starting 2008 June Network hardware redesign, new hardware to connect Effelsberg installed in Groningen 2009 Feb, Effelsberg changes ongoing Interstellar scattering probably a major problem Jupiter burst tests with Jean-Mathias Griessmeier Fringes between Exloo and Nancay Next major observation scheduled for Exloo Nancay--Effelsberg on 2009 Apr 06 2/31
Long Baseline Fringe Priority Long baseline testing given low priority while other more major issues attacked Getting beamformer to track the sky Clock tick offsets Antenna mode selection... I am busy with many other LOFAR responsibilities too We have two months in Effelsberg to prepare to install the high band antennas Limited long-baseline interest in Dwingeloo to push commissioning Everyone already busy with too many things LOFAR DiFX test observations in queue since 2008 Nov Jean-Mathias now pushing forward Jupiter science 3/31
So Why Am I Taking Up Your Time? LOFAR Long Baseline Working Group (LLBWG) Let the community know what we are up to Explain some of the calibration/imaging/processing challenges we are facing Show a few of the things we are doing to tackle our challenges Hopefully encourage general algorithm development for our specific needs I have at least two LOFAR postdoc positions in Bonn to fill New director, new interests --- old director, um, er, diversified interests Maybe you know of a good student or postdoc who could be enticed to move to Germany... 4/31
The Advantages of Moving to Germany Or, why did I move to Bonn? Dutch Version: Our former JIVE colleague James Anderson was recently visiting us. Already during his stay in Dwingeloo he had a well known love for fast German cars. His move to Germany was partially motivated by the cars that country produces, but also it's no speed limited highways. This picture shows one of his former colleagues drooling over his latest acquisition, with the proud owner standing next to it, showing off a machine that would certainly be restricted by the Dutch highway speed limits. --- Madroon E. J., Ranting My Version: The Bonn canteen is far, far better than the Dwingeloo canteen Cosmic symmetry resolution I moved to Socorro to do a PhD in astrophysics, and ended up briefly engaged to a woman who enjoyed doing Jackson E&M problems for fun. I moved to Bonn because my girlfriend lived in Cologne, and ended up becoming the LOFAR Project Manager for Bonn 5/31
LOFAR: The Low Frequency Array Aperture array technology Low Band (LBA) normally 30 to 80 MHz can do 10 to 80 MHz High Band (HBA) Core (2 km diameter) Remote (inside NL) International (outside NL) } digital processing 120 to 240 MHz 3rd input } Orginal LOFAR open at International stations extra LBA inputs for Dutch stations (better performance < 30 MHz) Current LOFAR 6/31
Core 2 km diameter Micky Mouse design Station Beam FWHM 8.7 6.6 30 75 120 240 MHz Synthesized beam 5.3 2.6 800 300 200 100 30 75 120 240 MHz 7/31
Remote Up to 130 km baselines Circular-pair half-design Station Beam FWHM 8.7 6.6 30 3.7 1.9 75 120 240 MHz Synthesized beam 20 30 8 5 3 75 120 240 MHz 8/31
International ~1000 km baselines Original station design Station Beam FWHM 9.9 4.0 30 1.7 0.7 0.4 0.2 30 75 120 240 MHz 8 International stations with reasonably secure funding 75 120 240 MHz Synthesized beam 2.5 1.2 Likely to be all that are available for the next several years 9/31
Why Are Long Baselines Important for Survey KSP LOFAR? Polarization calibration --- Anderson & de Bruyn both believe International stations necessary for LOFAR polarization calibration Resolution Reduce beam depolarization Investigate structure of polarization and RM changes on small spatial scales Cosmic Ray KSP Resolution --- sub-arcsecond resolution of HII regions in nearby galaxies, gravitational lenses, cluster structure, and so on Magnetism KSP Astrometry --- localization of faint, high redshift candidates to ~ arcsecond accuracy for optical follow-up Sensitivity --- International stations are the most sensitive Transient KSP Resolution --- localization of transients, study of AGN and Solar system 10/31
Technical Challenges for the LLBWG Science goals drive most imaging experiments to full field of view observations Time and bandwidth smearing limitations require small integration times and small frequency channels Different ionospheric conditions over distant stations require high time and frequency resolution to prevent subchannel/integration phase wrap Depending on ionosphere and frequency, can be more stringent requirement than time/bandwidth smearing 11/31
Data Rates Correlator polyphase filterbank generates 256 channels per subband by default 763 Hz channel width for 200 MHz clock 610 Hz channel width for 160 MHz clock Total of 63488 channels for standard 16 bit mode 253952 channels for 4 bit mode Visibility output rate limited to 50 Gb/s (6.25 GB/s) by current storage cluster input rate For LOFAR operation (18 Core, 18 Remote, 8 International stations): LOFAR Core (18 C stations) LOFAR Remote (18 C + 18 R stations) 16, 8, and 4 bit data modes all fit under storage rate limit 16, 8 bit data modes all fit under storage rate limit 4 bit fits for LBA, does not fit for HBA LOFAR International (18 C + 18 R + 8 I stations) All bit depths have storage rates too high 127 Gb/s for 16 bit mode, 508 Gb/s for 4 bit mode (15.9, 63.5 GB/s) 12/31
Time and Bandwidth Smearing 1000 km baseline LOFAR, International station beam, 120 MHz Correlator integration times in seconds, frequency averaging in terms of original 256 channels per subband 13/31
Smearing Commissioning Tests Can correct visibilities for time and bandwidth smearing Expensive software operation Small facets? Cannot regain lost sensitivity How well can this be done? Only works if reduction not too large (0.5? 0.8? 0.95?) Suppose we need dynamic range of 106 to 107, and we have a software fudge to correct a 10 or 20% smearing loss What about sources in far sidelobes? How well do we have to subtract them? 14/31
Image/Cube Sizes Certain science projects want to have full field of view available for analysis at once Sometimes we cannot process a single facet at a time, analyze, and throw away before moving to the next facet Rotation measure synthesis can lead to very large cube sizes Some projects such as Milky Way Galactic research, or nearby galaxy RM studies may require astronomer visualization of RM cube over large fraction of station beamsize, or even multiple beams 15/31
LOFAR σrm for Cosmic Magnetism RM error estimate shown as function of polarized signal and observing time Assumed spectral index α = -0.7 Observing time distributed to fill HBA frequency range Only get 0.1 rad m-2 in 10 s for a 10 mjy source Need ionospheric Faraday coherence time to be longer than 10 s 16/31
Observing the HBA Station Beam Station beam 2.5 to 1.2 degrees in size (120 to 240 MHz) Assume 2 degree beamsize for imaging LOFAR synthesized resolution ~ 0.4 to 0.2 arcseconds Assume typical 4 pixels across beam for imaging 1.1 x 1010 pixels per station beam (47 GB per frequency) RM resolution 0.1 rad m-2, RM range ±100 rad m-2 2.3 x 1013 pixels per RM cube (93 TB) Extragalactic RM search RM resolution 1 rad m-2, RM range ±10 000 rad m-2 2.3 x 1014 pixels per RM cube (930 TB) RM resolution 0.01 rad m-2, RM range ±10 000 rad m-2 2.3 x 1014 pixels per RM cube (93 PB) 17/31
LOFAR Long Baseline Calibration Commissioning First, get system to reliably produce fringes Deal with the ionosphere Deal with dipole/tile/station beam characteristics Commissioning Timeline 18/31
Vertical Total Electron Content Behavior Ionosphere 1 TECU = 1016 m-2 1 TECU 4/3 turn of phase at 1 GHz, or 40/3 turns at 100 MHz Ionization fraction lags Solar noon Electrons raised in equatorial fountain fall along flux lines to either side of equator 19/31
Crossing Points in Ionosphere Left: 1 beam with 3 4 sources More beams are extremely useful for 3-D tomography Right: 5 beams with 3 4 sources each Large amount of information at different heights New telescopes will see a calibrator source every 1 2 degrees over large part of the sky Equivalent to 3 7 km spacing at 200 km height in ionosphere LOFAR currently has capability to make up to 8 beams on the sky Cannot get crossing regions for widely separated stations 20/31
LOFAR and the Minimum Ionosphere Model (MIM) Work in progress to implement MIM in BBS (the standard pipeline software package for LOFAR) by M. Mevius and others Left: Westerbork LFFE data for 3C196 with obvious nonisoplanicity 21/31
Adding Terms to Model Ionosphere TL: 0th order polynomial TR: 2nd order polynomial BL: 3rd order polynomial 22/31
Long Baseline Ionosphere Should be enough flux using all sources in the beam to model ionospheric phase screen for individual International LOFAR stations Movie (Lenc et al. 2008) Real challenge will be getting ionospheric Faraday rotation on long baselines 23/31
Ionospheric Faraday Rotation Also Depends on the Earth's Magnetic Field Magnetic equator shifted toward Europe European antennas located at magnetic mid-latitudes Same for US, Australia, South Africa 24/31
Ionospheric Faraday Rotation Calibration Typical vertical daytime ionospheric Faraday rotation angle ~ 2 rad m-2 Solar minimum Polarized calibrator density (# per sq deg) for LOFAR too low to map out ionospheric Faraday rotation like we will do for delay Must rely on Faraday rotation models based on absoluteionospheric 3-D electron content Challenge is to derive absolute total electron content model from an interferometer which nominally only measures differences Roger Dean (Oxford) working with VLBI observations and GPSbased MIDAS (U Bath) ionosphere models to study problem 25/31
Beam Shapes and Long Baselines Dipole, tile, and station beam modeling important for proper calibration and high dynamic range imaging LOFAR stations have different layout orientations, so the beamshapes and sidelobe patterns will all be different Dipoles are supposed to be backrotated to allow minimal realtime polarization calibration and processing for tied-array modes Long baselines cannot have polarization orientations match for all directions on the sky Extra requirement for beam modeling to be implemented correctly 26/31
Antenna Beam Linear polarization dipoles Sensitivity strongly depends on azimuth and elevation angle Strong polarization leakage off-axis Analytic LBA dipole model by Yatawatta Because of high leakage terms, need to achieve effectively just as high dynamic range as Stokes I Don't win because true Q, U, and V are small Antennas fixed to ground, so nearly everything is off-axis Leakage typically 50% Little polarization sensitivity at low elevations Strong frequency dependence 27/31
Dipole Polarization Orientation Design considerations for LOFAR require dipoles to be aligned as well as possible for all stations HBA antennas have to be backrotated to correct for station layout orientation All stations end up with slightly different polarization beams Long baselines cannot be aligned perfectly for all directions on the sky Software calibration necessary 28/31
Timeline Now Work on getting fringes with standard network/correlator Develop and commission ionospheric calibration modes using multiple simultaneous station beams End of summer 2009 Should have 3 German stations plus several Dutch stations operating with HBA tiles (120 240 MHz) Search through 3C radio galaxies for sources with strong polarization to use for primary polarization calibrator grid Investigate absolute TEC measurement performance Start testing rapid switching mode for HBA tiles Tile beam can only see a small fraction of the sky, so must rapidly repoint tile beam around sky for some forms of polarization calibration Test time and frequency averaging possibilities for ionospheric and calibration performance 29/31
Timeline (Continued) Fall 2009 More International LOFAR stations scheduled to come on-line Hope to have all 8 International stations by end of year Initial real imaging tests More ionospheric tests as system sensitivity improves MS3 (initial all-sky commissioning survey) Long baselines used with the ~ 20 short Dutch baselines Data archived after initial RFI removal, preliminary calibration, and time/frequency averaging (total storage ~ 1 PB) Reprocessing to identify and measure medium-brightness polarized sources to create denser grid of polarization calibrators Simultaneously generate sub-arcsecond astrometric calibrator list Careful comparison with high-frequency source positions Position shifts with frequency and source morphology changes major issues 30/31
The End 31/31
Calibrators Polarized Galactic emission Limited to short baselines Extremely complicated rotation measure features Pulsars Many pulsars are strongly polarized at low frequencies Interstellar scattering probably limits baselines to less than a few hundred km for pulsars in the Galactic plane 10 s ionospheric calibration timescale difficult to achieve with pulsars with pulse periods ~ 1 s (pulse to pulse variations) AGN jets and lobes Probably require long baselines (> 100 km) to minimize beam depolarization 32/31
Fan Region (Galactic Emission) Westerbork LFFE observations Strong polarized emission Complicated RM structure 33/31
Pulsar Observations with LOFAR Many pulsars show strong polarization at low frequencies Averaging over pulse phase reduces fractional polarization --- may need to correlate with pulsar binning Pulsars already detected with LOFAR Westerbork LFFE pulsar search planned soon 34/31
AGN Polarization Many high brightness sources should have substantial polarization Expect significant polarization intensity out to long baselines Also important for use as probes of magnetic fields in other objects Important for International station calibration At LOFAR sensitivity, there should be a relatively high surface density of calibrators Improves ionospheric mapping, but not sufficient 35/31