KAT-7/MeerKAT Commissioning/ Science operations
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
KAT-7/MeerKAT Commissioning/ Science operations SPT and Commissioning teams * http://public.ska.ac.za/kat-7/kat-7-operations * Bennett, T., Blose, S., Booth, R., de Villiers, M., Dikgale, A., Foley, T., Frank, B., Goedhart, S., Hess, K., Horrell, J., Lucero, D., Magnus, L., Mauch, T., Nemalili, T., Oozeer, N., Passmoor, S., Ratcliffe, S., Richter, L., Schwardt, L., Smirnov, O., Spann, R., Williams, L., Wilson, S., Wolleben, M., Zwart, J.,
Operational nodes • Cape town o Full primary operations conducted on a service observation basis to ensure little or no time on site • Klerefontein o Maintenance primary node used to monitor and plan corrective and preventative maintenance • Losberg o Primary node during commissioning but will allow full operations if needed as well an access point for the deployment of user installed hardware
Operational risks identified • Staff at remote sites tend feel isolated • Observers lose motivation if they don’t visit the telescopes • There is a decrease in radio astronomy expertise within the broad user community due to less exposure to the telescopes • Less feedback between site and operator staff leads to increase in faults not being quickly resolved • Inadequate information flow leads to a loss of “latest news” information to observers. • Fewer observers at site leads to poorer communication with engineers and ultimately a drop in science quality. • Scheduling changes on site make it difficult to support some types of observing programs.
Operational risks identified • Staff at remote sites tend feel isolated • Observers lose motivation if they don’t visit the telescopes • There is a decrease in radio astronomy expertise within the broad user community due to less exposure to the telescopes • Less feedback between site and operator staff leads to increase in faults not being quickly resolved • Inadequate information flow leads to a loss of “latest news” information to observers. • Fewer observers at site leads to poorer communication with engineers and ultimately a drop in science quality. • Scheduling changes on site make it difficult to support some types of observing programs.
MeerKAT rollout • 2013 - First qualification dishes deployed • 2014 - Install test and commission 6 antennas using a dedicated test system • 2015 - first official correlator support • 2015 - 29 antenna at about 3 per month • 2016 - 29 in 8 months at about 4 per month • Basic rule ... don't' panic
Single dish • Surface accuracy • Spectral modes • Pointing calibration • Primary beam • Gain calibration • Polarisation • Search for unpolarized and polarized, but stable standard sources • RFI • System linearity
KAT-7 – RFI Flagging
KAT-7 – RFI Flagging
Single dish • Surface accuracy • Spectral modes • Pointing calibration • Primary beam • Gain calibration • Polarisation • Search for unpolarized and polarized, but stable standard sources • RFI • System linearity
Single dish • Surface accuracy • Spectral modes • Pointing calibration • Primary beam • Gain calibration • Polarisation • Search for unpolarized and polarized, but stable standard sources • RFI • System linearity
Single dish • Surface accuracy • Spectral modes • Pointing calibration • Primary beam • Gain calibration • Polarisation • Search for unpolarized and polarized, but stable standard sources • RFI • System linearity
Single dish • Surface accuracy • Spectral modes • Pointing calibration • Primary beam • Gain calibration • Polarisation • Search for unpolarized and polarized, but stable standard sources • RFI • System linearity
Single dish • Surface accuracy • Spectral modes • Pointing calibration • Primary beam • Gain calibration • Polarisation • Search for unpolarized and polarized, but stable standard sources • RFI • System linearity
Single dish • Surface accuracy • Spectral modes • Pointing calibration 0.026 (0.001) • Primary beam • Gain calibration 63.9 (3.8) • Polarisation HH 24.2 (1.4) • Search for unpolarized and polarized, but VV 24.4 (1.2) stable standard sources 938 (65) • RFI • System linearity
Some problems …
Some problems …
Imaging array • Dynamic fringes • Antenna location (delay) calibration • Phase Closure • Measure cross-correlation phase stability • Amplitude closure • Measure interferometric gain stability • Baseline calibration • Baseline specific errors • Band pass calibration and stability • Absolute flux calibration • Interferometric polarisation and calibration
Imaging array • Dynamic fringes • Antenna location (delay) calibration • Phase Closure • Measure cross-correlation phase stability • Amplitude closure • Measure interferometric gain stability • Baseline calibration • Baseline specific errors • Band pass calibration and stability • Absolute flux calibration • Interferometric polarisation and calibration
Imaging array • Dynamic fringes • Antenna location (delay) calibration • Phase Closure • Measure cross-correlation phase stability • Amplitude closure • Measure interferometric gain stability • Baseline calibration • Baseline specific errors • Band pass calibration and stability • Absolute flux calibration • Interferometric polarisation and calibration
Imaging array
Primary beam correc,on 20 arcmins Primary beam correction was performed in MIRIAD using a 1.3 degree Gaussian (at 1.4 GHz). So far primary beam correction has not been confirmed in CASA.
Confirma,on of beam center polariza,on calibra,on • Polarization calibration: verification using 3C286 and 3C138 with Miriad o 12 hour observation, first 6 hours 3C138, then 3C286 o Used 3C286 as gain and polarization calibrator o Bootstrapped calibration to 3C138 (both sources are not visible at the same time) o Miriad flux models for these sources at 1.4 GHz: § 3c286: IQUV = 14.6, 0.56, 1.26, 0.0 (flux calibrator) § 3c138: IQUV = 8.6, 0.60, -0.27, 0.0 (test source) o Observed fluxes at 1.4 GHz: § 3c286: IQUV = 14.6, 0.56, 1.26, 0.0 as expected § 3c138: IQUV = 8.5, 0.58, -0.26, 0.0 calibration confirmed! ☺ o Conclusion is that Miriad works very well with linear hands of polarization and there is nothing, at this point, that prevents us from doing polarization imaging once the online system is there.
Polarization Observa(on: off axis performance • 3C286, 5x5 raster, spaced 20 arcmin, cycled through every 30 minutes • observed from source rise ,ll set to get good parallac,c angle coverage • the centre poin,ng was used for calibra,on (Miriad, full polariza,on) • each 3C286 poin,ng imaged individually and then source from each image copied into a single fits file to produce the images below Recommenda(on: • The useful FOV is
Polarized intensity 3C286 The degree of polarization is given by √(Q2 + U2 + V2)/ I2 The image (left) confirms that the Mueller matrix is position-dependent and variable across the beam. 3C286 has the following polarization parameters I Q U V 14.8 0.56 1.25 0
Q and U across the beam 3C286 Work is underway to model this variation so as to be able to use a greater portion of the beam for polarimetry
Amount of circular polariza,on 3C286 There is no circular polarization at the centre of the beam (expected) but there appears to be leakage off beam centre. Calibration of this is ongoing together with a scheme to do L-Band holography for a measurement of beam characteristics.
Science verification • Track commissioning parameters with time • Imaging tests with a range of spectral resolutions, source complexity, brightness, in single-field interferometric mode, Dynamic range, Reproducibility, Noise level (done in CASA and MIRIAD) • Mosaicing • Polarization: done on beam center not full beam • Fine tune observation procedures • Calibrator surveys • Full beam calibration (Meqtrees) • Spectral mode imaging, beamformer and VLBI still to be confirmed
Science verification • Track commissioning parameters with time • Imaging tests with a range of spectral resolutions, source complexity, brightness, in single-field interferometric mode, Dynamic range, Reproducibility, Noise level (done in CASA and MIRIAD) • Mosaicing • Polarization: done on beam center not full beam • Fine tune observation procedures • Calibrator surveys • Full beam calibration (Meqtrees) • Spectral mode imaging, beamformer and VLBI still to be confirmed
Milestones for May $3000 raw voltage capture machine 35
Pulsar Timing
Science verification overview • Mosaicing / polarization - Maik Wolleben o Mosaicing observation planning and reduction confirmed o Primary beam correction • Spectral lines - Sharmila Goedhart, Brad Frank o NGC 3109 HI o HI absorption o MASERS • Continuum observations - Nadeem Oozeer o El Gordo • Target of opportunity - Tony Foley (TAC) o ATEL #3694
The next three slides show the work done to firstly plan mosaicing observations and second to reduce the data with the required calibration and corrections applied. Mosaicing • Left: KAT-7 Mosaic consisting of 49 Fields, 1.9 GHz • 10 hours total observing time (including calibrators) • Right: NVSS at 1.4 GHz shown for comparison
Mosaicing • 1225 fields • observed during 3 consecutive nights • size: 10 x 10 deg in Galactic coordinates • integration time of each field: 30 s • RMS in the final image: 35 mJy/beam in Stokes I, 20 mJy/beam in polarized intensity. Probably due to ground radiation. Supernova Remnant SN 1006. (Type 1A observed in 1006 AD.)
Spectral modes Mode Band Bandwidth Channel Bandwidth Available Wideband 256 MHz 390.625 kHz Yes 54000 km/s 82 km/s 8k Wideband 256 MHz 48.8 kHz Yes 54000 km/s 10 km/s OH Spectral Line 6.25 MHz 1.5 kHz In test (Search) 1300 km/s 0.3 km/s OH Spectral Line 1.5 MHz 381 Hz In test (Monitor) 327 km/s 80 m/s HI Spectral Line 33.3 MHz 4 kHz ~ Oct 2012 7000 km/s 0.84 km/s Velocities are referenced to HI http://public.ska.ac.za/kat-7
HI Velocity Field of NGC 3109 Observed using the wbc8k mode (10km/s spectral resolution for HI). The wbc8k mode was made available to commissioning to start testing observation and reduction strategies for the narrow band modes while the other modes are being developed. It illustrates the power of the ROACH design as the time from conception to first deployment was one day.
PKS1814-637 absorption This was a short (10 min on source) observation to confirm the presence of the absorption structure The expected absorption line is at 1334 MHz.
MASER line for NGC6334 in mode c16n2M4k
MASER lines for NGC6334 in mode c16n7M4k
Images (line maps) at 1665 and 1667 MHz respec,vely
Overlay of the narrow band contours on the con,nuum image
El Gordo (ACT-CL J0102-4915) Aim for observation • check for diffuse synchrotron emission • check polarization • derive spectral index and look for break in spectrum (derive/ confirm physical parameters) • if no detection: may get only an upper limit • check expected rms using long integration.
Comparison with SUMMS These sources are just above the detection limit of SUMSS but well detected with KAT-7
PKS 1510-089 - ATEL-3694 Appart from monitoring the source PKS 1510-089 for the ATEL, the regular observations have been used to test the auto-delay (fringe stopping) functionality and to test interferometric pointing.
Circinus X-1 – Observations of most recent flare show good radio brightening 51
Summary • The industrial commissioning plan is being fine tuned • The software tools are in place to characterize the system • The operations and commissioning teams are getting valuable experience in using KAT-7 • We are already doing some continuum/ transient science • Plan for operations …
You can also read