Title: Using masers as evolutionary probes in the G333 GMC (as well as some follow up work)
1Using masers as evolutionary probes in the G333
GMC (as well as some follow up work)
- Shari Breen, Simon Ellingsen, Ben Lewis, Melanie
Johnston-Hollitt, Simon Wotherspoon, Indra Bains,
Michael Burton, Maria Cunningham, Nadia Lo Tony
Wong
2Talk Outline
- 1 - The G333 GMC
- The region
- Our observations
- Analysis and results
- 2 - Follow up observations and analysis with 1.2
mm dust clumps from Hill et al. (2005) - Water masers
- 12 GHz methanol masers
3Masers and mm mapping data
- Masers pinpoint special sources within mm
mapping data - mm mapping data along with maser observations may
hold the key to unlocking the evolutionary stages
that the different species of masers are tracing
4The G333.2-0.6 GMC
- Located at 3.6 kpc (Lockman 1979)
- Extends 1.2 x 0.6 degrees on the sky
- 61 13CO (Bains et al. 2006) clumps
- 95 1.2 mm dust continuum
- clumps (Mookerjea et al. 2004)
- Completely searched for
- 6.7 GHz methanol masers
- (Ellingsen et al. 1996) and
- 1665 and 1667 MHz OH
- masers (Caswell et al. 1980)
5Observations
- Mt Pleasant
- Two regions completely searched in 2005
- Velocity coverage of 431 km s-1 centred on -40 km
s-1 - Region 1 750 pointings
- RMS noise limits below 1 Jy in 78
- of data
- Region 2 300 pointings
- RMS noise limits below 1 Jy in 35
- of data
- 9 water maser detections, 5 of
- which are new discoveries
6Observations
- ATCA
- Two sessions of ATCA Directors time
- Preliminary observations
- 2006 June 19, 1.5D array, 8 masers detected
- Final observations
- 8 masers observed on 2006 July 16 17, 6A array
- RMS noise limits 0.15 Jy in a single spectral
channel - RMS noise in continuum images typically 0.02 Jy
beam-1 - Positions accurate to about 0.5 arcseconds
7Water masers - circles Methanol masers - squares
(Ellingsen 2005) OH masers - crosses (Caswell
1998)
84
- Red 8 µm, green 5.8 µm, blue 3.6 µm. Water
masers (circles), methanol - masers (squares) (Ellingsen 2005) and OH masers
(crosses) (Caswell 1998).
9Associations between maser species and IR sources
- More water (9) masers than either methanol (4) or
OH (3) - Relatively few associations between species
- Water masers lie along the main axis of star
formation, methanol tend to be located near the
periphery - 4 of the 5 new detections have associated GLIMPSE
point sources - 4 previous detections are clearly projected
against regions of bright mid-IR emission
10Associations between maser species and IR sources
- Comparison of the 3.6-4.5 colours of water
maser associated GLIMPSE sources with those
associated with methanol masers (Ellingsen 2006)
shows that the water associated sources are
clustered towards the less red end of the range
observed in the methanol associated sources - Methanol masers trace a generally earlier
evolutionary phase than water masers
1113CO clump analysis Water Masers
- Increasing probability of maser presence
associated with increasing values of all clump
properties (confidence 95 or higher) - Estimated regression relation is
- Misclassification rates are good when probability
threshold set to 0.5 - 33 clumps in region with no associated maser -
model predicts 31 correctly - 7 clumps with associated masers - model predicts
5 of these have associated water masers
121.2 mm clump analysis Water Masers
- Increasing probability of maser presence
associated with increasing values of all clump
properties (95 confidence for higher) - Estimated regression relation is
- Misclassification rates are good when probability
threshold set to 0.5 - 66 clumps in region with no associated maser -
model predicts 65 correctly - 7 clumps with associated masers - model predicts
4 of these have associated water masers
13Physical implications of the dust/water maser
model
- Model only uses radius as a predictor
- No more information is offered from any other
property - If set pi to be 0.5 corresponds to a clump radius
of 1.25 pc
Clumps with r 1.25 pc have a probability of 0.5
or greater of having an associated water maser
1413CO clump analysis Methanol Masers
- Increasing probability of maser presence
associated with increasing values of column
density and integrated flux - Estimated regression relation is
- If pi is set to 0.5 the model is good at
determining clumps with no associated maser - If pi is set to 0.12 the model correctly predicts
the 5 clumps with associated masers and falsely
predicts that 9 of the 56 clumps with no
associated maser emission
151.2 mm clump analysis Methanol Masers
- Increasing probability associated with clump
density, radius and mass - Estimated regression relation is
- Good misclassification rates when probability
set to 0.5 - 90 dust clumps have no associated maser model
predicts correctly for 89 - 5 clumps with associated masers model correctly
predicts 3 of these
16Conclusions from the G333 observations
- More water masers than either methanol or OH
masers, with few associations between species - Water masers located close to the main axis of
star formation within the GMC - Water masers associated GLIMPSE point sources are
similar to those associated with methanol masers
although slightly less red - Water masers are associated with bigger,
brighter, more massive and denser 13CO and 1.2 mm
dust clumps - Present models with low misclassification rate
that can be tested on larger scales
17Testing the dust/water maser model
- Hill et al. (2005) observed 404 1.2 mm dust
clumps - 70 clumps have a probability of 0.5 or greater
of having an associated water maser - Preliminary results have shown that more that 75
of the masers that we detect are associated with
clumps that we predicted had a high probability
of maser presence - Will complete ATCA observations in 2008 August
1822 GHz results
- Majority of water masers lie in the centre of the
1.2 mm dust clumps - No 22 GHz continuum detected that is not
associated with a dust clump that is also
associated with a water maser
19(No Transcript)
20Methanol masers at 6 12 GHz
- Lewis (2007) targeted observations of 12 GHz
methanol masers towards 6.7 GHz methanol masers - 27 target sources overlap with Hill et al. (2005)
1.2 mm dust clump sources - 12 GHz masers detected at 17
- Only clump mass can be used singularly as a
predictor - Implies that 12 GHz methanol masers are
associated with less dense clumps than 6.7 GHz
methanol only clumps
218
22Methanol masers at 6 12 GHz
- Estimated regression relation is
- Model correctly predicts 16 of the 17 clumps with
associated 12 GHz masers and 6 of the 6.7 GHz
methanol masers without 12 GHz emission
12 GHz methanol maser associated with a later
evolutionary phase of massive star formation than
the 6.7 GHz only sources
23Summary of main results
- 6.7 GHz methanol masers trace the earliest stages
of massive star formation followed by 12 GHz
methanol masers - supported by theory - water masers present at a later evolutionary
stage
2413CO and 1.2 mm dust clump analysis Water Masers
- Fitted a Binomial GLM to maser presence/absence
data - Predicts the probability pi of finding a water
maser in the ith clump in terms of clump
properties x1i, x2i,.xmi - Where yi is the maser presence or absence in the
ith clump and b terms are the regression
coefficients to be estimated - stepwise model selection used to select the
simplest model with the greatest predictive power