eMERLIN

1
e-MERLIN Radio issuesSimon Garrington

• Wide-fields (ie 10x10 arcmin)
• uJy sensitivity 50-150 mas resolution key
  science driver for e-MERLIN
• 1500 sources for deep 1.5 GHz field
• AG, VO tools already essential for multi-l
• Making transition from expert to routine use

2
e-MERLIN Project Status

• DESIGN 2x2 GHzwidefieldfreq. flexible
• Front-ends
• New C-band (4-8 GHz) installed
• L-band lens, mechanisms, horns in place
• IF,samplers
• Design/prototype on telescope 05-06
• Fibre network
• 90 km new dig 600km trunk
• All installed connected
• First light on local sections
• Cambridge-Birmingham tests start today
• Digital transmission
• Modules ready for tests this year
• Correlator
• WIDAR from DRAO
• 16 x 128 MHz sub-bands 2048 complex lags
• 1st fringes prototype 2006 (4 tels 1 band)
• Full correlator 2007/8

3
e-MERLINsoftware/imaging challenges

• All 1.5 GHz (5GHz) images will require wide-field
  mapping
• To reach 1-3uJy noise in normal 12-24hr runs
• Wide-field wide-band imaging techniques
  algorithms
• EVLA, LOFAR, WSRT,LWA ? KAT, NTD ? SKA
• Image-plane calibration (ionosphere antenna
  pointing)
• Wide-band imaging MFS
• Dwingeloo workshop June 05 www.skatelescope.org
• ALBUS (RadioNET JRA) www.radionet-eu.org
• Advanced Long Baseline User Software
• ASTRON,JIVE,MPIfR,JBO
• Parallelisationcalibrationwide-fieldwide-band

4
ALBUS Progress

• Smarter pipelines/scripts ... Improving the
  products delivered to the users/archives
• Smart calibration
• Using GPSIonospheric models (Anderson. JIVE)
• Using WVR/GPS for tropospheric calibration (Roy,
  MPIfR)
• Better pipelines
• ParselTongue Python for AIPS ideal for CEA
  integration
• (Kettenis, JIVE)
• Wide-band Wide-field
• data splitting task parallelisation
  (JIVE, JBO)
• Opportunity for AG Workflow management

5
Summary

• New generation of cm- and metre-wave radio
  telescopes in 2007gt
• New ways to access data
• Some (eg LOFAR) will be radically different
• Others eg e-MERLIN,EVLA
• Lots more data wider fields new algorithms
  required faster smarter pipelines
• Expectation that data may not move
• Increased sensitivity will motivate more multi-?
  studies using VO tools
• Of course, data will be VO-ready
• Renewed interest effort in algorithms and
  processing strategies
• No big new packages interfaces to legacy
  code workflows etc. ParselTongue opportunity