A.R. Offringa - LOFAR

Post on 07-Jan-2022

3 views 0 download

transcript

LOFAR

André Offringa – June 15, 2011

The LOFAR-EoR project

and the LOFAR collaboration

(many slides from Ger de Bruyn and Michiel Brentjes)

LOFAR(and a little bit of RFI)André Offringa – June 15, 2011

The LOFAR-EoR project

and the LOFAR collaboration

(many slides from Ger de Bruyn and Michiel Brentjes)

The LOFAR telescope

LOFAR's key properties:● 10-90 and 110-300 MHz● Beamformed stations from dipoles, tiles● Wide field of view● Very flexible observing modes● Large collecting area● Core in the Netherlands● International baselines up to 1000 km.

The low and high band antennas

Beam former

RS208

RS503

RS307

CS302

RS106

WSRT

18 core stations(all split-stations with 2x24 tiles)

+ 4 more stations in core(CS011,013,028,031)

1 km

Superterp core stations

Core stations

HBA tiles:2x24LBA dipoles:96

Effelsberg

Tautenburg 96 tiles

International stations

~20 deg

~100

deg

digitally formed station beam

Inner/outer 48 dipoles

Widefield beam for HBA & LBA

Stations are calibrated

● Using redundancy :)

LOFAR observing modes

LOFAR's correlator

● Software correlator (Blue Gene, Groningen)● Very flexible● Superb frequency resolution: 0.76 kHz● Default integration time: 1 sec● Currently 48 MHz bandwidth, upgrade to 96

(192?) MHz.● Bandwidth and beams interchangeable, e.g.

24 x 2 MHz beams.

LOFAR key science projects

● Epoch of Reionization!● Pulsars & transients● Survey● Magnetism● Solar (system)● Ultra high energy physics / CR

The LOFAR – EoR project

● Exposure/field: 100 x 6h● 112-190 MHz● Raw resolution: 1s / 0.76 Khz● After RFI detection, average time and freq

5-10 times.● Raw data rate: 30 TB/h● Averaged data rate: 0.2 – 0.4 TB/h

EoR fields

LOFAR imaging results

3C196 (Wucknitz)

Cas A (Yatawatta)

Cas A (Chandra)

LOFAR offline processing

● Standard pipeline:– Flags (AOFlagger),

– Averages (NDPPP),

– Subtracts bright sources (demixing),

– Calibrates (BBS),

– Images

LOFAR offline processing

● Offline processing on cluster of ~100 nodes

● Sub-bands are concurrently processed.● Averaging + flagging takes 25%-50% of

observation time.

AOFlagger

● Very fast & accurate

● Rough estimate of astronomical data

● SumThreshold● Iterative (2.5x)● Density dilation

LOFAR and RFI

● RFI environment behaves well● Few percentage of data requires flagging● High frequency resolution!● Antennas on the ground

LBA Total power, before flagging

LBA Virgo, total power, after flagging

HBA “RFI survey” observation, percentages

RFI survey: variance of channel difference (“the noise”)

RFI

● Possible 2nd stage flagger● Same algorithm, other input

LBA RFI

HBA RFI

4 observations combined

Summary

● LOFAR works● RFI environment is benign