Milliarcsecond Morphology of Scintillating

1
Milliarcsecond Morphology of Scintillating
Non-scintillating Sources
Roopesh Ojha, Alan Fey, David Jauncey, Jim
Lovell, Ken Johnston
Email rojha_at_atnf.csiro.au
Dwingeloo, April 7th, 2004
2
Outline
Background Motivation-Objectives The three
samples The observations Analysis Results Conclusi
ons
3
The MASIV Survey
The Micro Arcsecond Scintillation Induced
Variability (MASIV) Survey, aims to construct a
sample of 100 to 150 scintillating extragalactic
source with which to examine both the
microarcsecond structure and the parent
populations of these sources, and to probe the
turbulent ISM responsible for the scintillation.
VLA observations have been made, at 4.9 GHz, of
710 compact, flat-spectrum sources over the
Northern sky. We have observed four epochs spread
evenly over 13 months. The first epoch revealed
variability on timescales ranging from hours to
days in 85 of the sources. The number of highly
variable sources, those with rms flux densities
above 4 , increases with decreasing source flux
density but rapid, large amplitude variables such
as J18193845 are very rare.
Dwingeloo, April 6th, 2004
4
Motivation
Increase in fraction of highly variable
scintillators (those with rms flux density
variations above 4) with decreasing flux
density Increase in their fractional amplitude
of variability with decreasing flux
density These results of the MASIV survey raised
the possibility that the Low Flux Density
Scintillating sources may form a distinct
population with differing morphology
5
Milliarcsecond Scale Structure of Scintillating
Sources
Our VLBI observations address the following
questions

• Are there any morphological differences, at mas
  scales, between high flux density, 1 Jy, and
  low flux density, 0.1 Jy, scintillating sources
  ?
• Are there any morphological differences, at mas
  scales, between Scintillating and
  Non-scintillating sources ?

Dwingeloo, April 7th, 2004
6
The Observations

• Used the VLBA to make 8.4 GHz snapshot images of
  the 75 Low Flux Density Scintillating (LFDS)
  sources found with the MASIV survey.
• 4 scans of 7.5 minutes each, full polarization
• Usual AIPS, Difmap, Model fitting.
• Mean and median flux density of 120mJy and
  110mJy
• Distribution of fitted core angular size (core
  assumed to be fitted component at origin of
  image) has mean (median) of 0.11 (0.10) mas
• 19 sources have completely unresolved cores
• Mean separation from the core 2.9 mas

7
Fig 1 Examples of LDSS sources
Dwingeloo, April 7th, 2004
8
The Three Samples
Three uniformly selected samples in that

• Point sources at all resolutions of the VLA
• Spectral index flatter than 0.5 (from 1.4 to 4.9
  GHz)
• 8.4 GHz with VLBA
• Two Scintillating samples were found to be
  scintillating with rms flux density variations
  during a 72 hour period of
• S rms gt ((0.0032 (0.02S)2))

9
LFDS
LOW FLUX DENSITY SCINTILLATING Source flux
density ranges from 50 to 370 mJy with mean of
120 mJy 75 sources Our observations
10
HFDS
HIGH FLUX DENSITY SCINTILLATING Source flux
density 1 Jy 18 sources Structural information
from USNOs, RRFID. Database of images of all
radio reference frame sources at the same
wavelengths as those used for precision
astrometry. To monitor sources for variability or
structural change.
11
HFDN
HIGH FLUX DENSITY NON-SCINTILLATING No
scintillating behaviour detected by MASIV Source
flux density 1 Jy 144 sources Structural
information from USNOs, RRFID.
SEA ?
12
Dwingeloo, April 7th, 2004
13
Analysis

• Used three measures
• Core Fraction (core dominance)
• Flux-weighted radial extent (core dominance)
• Unweighted radial extent (overall source size)

Dwingeloo, April 7th, 2004
14
Core Fraction
Ratio of core flux density to total flux
density C Si (Si)beam / Si Si Core flux
density is defined as the sum of the CLEAN-ed
flux density within one synthesized beam Total
flux density is defined as the total CLEAN-ed
flux density (i.e. the sum of all CLEAN
components)
Dwingeloo, April 7th, 2004
15
LFDS(HFDS) and HFDN 99 (95) level LFDS-HFDS
same pop 10
Dwingeloo, April 7th, 2004
16
Flux-weighted radial extent
Complementary measure of core dominance, which is
defined as R Si Siri / Si Si Where ri is the
radius at which the ith CLEAN component has flux
density Si R has units of milliarcseconds
Dwingeloo, April 7th, 2004
17
LFDS(HFDS) and HFDN 99 (99) level LFDS-HFDS
same pop 68
Dwingeloo, April 6th, 2004
18
Unweighted radial extent
Straightforward measure of overall source size
i.e. the maximum radial extent of the source. It
is the angular radius that contains 95 of the
total CLEAN-ed flux density. We first calculate
a quantity E, similar to C above E Si
(Si)? / Si Si Where the numerator represents the
flux density contained within an area of angular
radius ?.
Dwingeloo, April 7th, 2004
19
LFDS(HFDS) and HFDN 99 (97) level LFDS-HFDS
same pop 50
Dwingeloo, April 6th, 2004
20
Caveats

• HFDS sample smallish (18) --- but large enough
  for K-S test to be valid
• Comparison of LFDS sources with HFDN sources
  sensitivity limits could bias the results as weak
  extended structure in the LFDS sample might not
  be detected.

Dwingeloo, April 7th, 2004
21
Results of Analysis
Both low and high flux density scintillator
samples have significantly higher core-fraction
and significantly smaller flux-weighted linear
extent than the non-scintillating sample ?
significantly more core dominant Overall angular
size of scintillating sources is significantly
smaller than that of non-scintillators

Dwingeloo, April 7th, 2004
22
Results of Analysis
There are no significant differences between the
mas morphology of low and high flux density
scintillators Consistent with same parent
core-fraction population

Dwingeloo, April 7th, 2004
23
Conclusions
Both low and high flux density scintillators have
significantly different morphologies than
non-scintillators have a higher
proportion of their flux in a compact core
have a smaller overall angular size. Low and
high flux density scintillators do not have
significantly different morphologies. ? Cores
scintillate LFDN sample and prediction
Dwingeloo, April 7th, 2004
24
Statistics of Source Angular Extent
C
R(mas)
q0.95(mas)
Dwingeloo, April 7th, 2004
25
K-S Statistics
K-S probability
Dwingeloo, April 7th, 2004
26
ABSTRACT
We present VLBA snap-shot images of low flux
density scintillating sources and compare their
morphologies with those of high flux density
scintillating sources imaged by the USNO
astrometry program. It is now clear that
interstellar scintillation (ISS) is the principal
cause of the intra-day variability (IDV) seen in
some compact, flat-spectrum radio sources at
centimeter wavelengths. The MASIV
(Micro-Arcsecond Scintillation-Induced
Variability) survey has been looking for new
scintillating radio sources. One result from this
survey is that the fraction of scintillating
sources increases strongly with decreasing flux
density. Here we address the possibility that the
above MASIV result arises because we are seeing
different mas morphologies among the weaker
sources.
Dwingeloo, April 6th, 2004