The annual science meeting was held at l'université de
Montréal on May 12th and 13th. For those who could not attend,
below is the meeting's programme. Some of the talk titles are linked
to abstracts or articles which summarize the presentation. When these
were not supplied, a "mailto" link to the presentor's name is provided
so that you can easily write for more information.
A few visitors came to talk to us
about other surveys which may be relevant to future CGPS plans.
I left my heart in Montreal.
The annual CGPS science meetings get better each year. Over 40 colleagues attended this year's meeting in Montreal. Thanks Claude and company for hosting the meeting, and providing such a splendid atmosphere. The management committee will long remember the meetings at the book store, and our warm interactions with the locals.
The full gamut of science topics, from shells and SNRs to HII regions, cold clouds and dark arcs, was covered. At a separate polarization workshop discussions with our colleagues in Germany and the US have lead to ambitious plans for collaborations, and highlighted some very intriguing puzzles related to the polarized radiation from the Galaxy. If we are successful in launching our phase II project, we can look forward to many more annual gatherings.
Significant progress was made in phase II planning at the meeting. We have now converged on a "battleship" geometry for the phase II observations, with a full longitude extension over a range of 65 to 180 degrees forming the hull of the boat and a high latitude section forming the bridge. The high latitude component will allow studies of the disk-halo interaction up to several kpc, HI imaging of the local Cepheus star formation region, and studies of the magnetic-ionic Faraday screen to high latitudes. A letter of intent for the NSERC Collaborative Research Opportunities program will be submitted by the end of June. The subsequent proposal will request fund to support data processing and University science involvement in the phase II project.
Part of the phase II project is the Global HI Conspiracy to image almost the entire Milky Way disk in HI by combining data from the CGPS, SGPS and VGPS. The good news between this newsletter and the last is that the VGPS has been awarded an initial allocation of 260 hours in the upcoming D-array to survey the section from 67 to 18 degrees longitude. Pilot observations for the survey took place shortly after the Montreal meeting on May 18. Those data have now arrived in Calgary and will be processed in the very near future. The first VGPS proper observations are scheduled for July 22. Observations then proceed at a very rapid pace during the last part of July and first part of August. It will be a very busy September.
The DRAO 9m array antennas have been modeled in order to know more about instrumental polarization characterizing these antennas at the radio continuum frequency of 1420 MHz using Antenna Theory. Until now, experimental results obtained by Peracaula (1999) have been used to correct observations within 75' from field center with a fair level of confidence. Theoretical calculations using the reflector antenna modeling software Grasp8 were undertaken to better understand the instrumental polarization within this 75' and extend this zone of confidence up to 90' in order to increase the usable field of view of the polarimeter and help in mosaicing the images.
Preliminary results look promising. There seems to be some agreement between theoretical and experimental data, especially near the field center. One element that has a strong impact on the results is the number of feed-supporting struts possessed by an antenna. The 3-strut model agrees more with the experimental results than the 4-strut model. This is not surprising however, since the array is dominated by 3-strut antennas. One important improvement would be to `average' the 3- and 4-strut results so we could directly compare theoretical results with experimental results since the ultimate goal is to simulate the whole array. In fact, this step is mandatory if one wants to use Grasp8's results to correct future observations. A complete report on this work will be published in the 2000 CGPS polarization workshop proceedings.
The 7C(G) survey covers much of the northern Galactic Plane Survey at 151 MHz, with a resolution of about 1.2 x 1.2 cosec(dec) arcmin**2 (EW x NS). The ``G'' stands for Galactic, to distinguish this survey from other 7C surveys away from the Galactic plane that have also been made with the Cambridge Low Frequency Synthesis Telescope (CLFST). Images from this survey, both at the full resolution, and at a resolution of about 4 x 4 cosec(dec) arcmin**2 (EW x NS), are now available on the MRAO website, see:
These images have been regridded to a B1950.0 grid, and corrected for the primary beam of the CLFST. Details of the observations, data reduction and calibration of this survey are described in Vessey & Green (1998, MNRAS, 294, 607).
I have taken some steps to improve the FCRAO CO(J=1-0) Outer Galaxy Survey. The initial release of the FCRAO OGS (Heyer et al 1998 ApJS 115,241), subsequently incorporated into the CGPS, contains several residual artifacts. These are :
The baseline ripple arises from incomplete cancellation of a standing wave between the subreflector and dewar at FCRAO. While an attempt was made to suppress this in the first release, this was not completely successful, due to insufficient sensitivity in any single spectrum. However, the ripple is repeated in similar form in spectra taken closely spaced in time, and so a better approximation to the ripple is possible.
The OGS is position-switched data, and any single reference position is generally shared over spatial scales of around a degree or so. This means that multiple independent measurements of the reference position are present in the OGS. It is therefore possible to exploit this redundancy and obtain a sensitive estimate of the CO(1-0) in the reference position and remove it from the data. The existence of reference contamination is itself interesting as according to theory we should have none at our sensitivity level.
The recurrent noise induced by the sharing of reference measurements is a more entrenched artifact. A total of 10 'on source' measurements share the same 'off source' measurement. However, since the positions of the spectra that share a reference measurement are not all spatially contiguous some steps can be taken to suppress the noise via deconvolution algorithms. (This recurrent noise is the origin of the 'cheesegrater' appearance of the FCRAO OGS.)
The baseline ripple and reference contamination preclude the construction of a sensitive account of the CO(1-0) emission based on large-scale summation/averaging, while the noise recurrence plays havoc with cloud-finding schemes (if you get one false cloud, you typically get 4 more false ones along with it).
I have made a first-pass test of a reprocessing to suppress/remove these artifacts. This has (1) flattened the baselines, (2) removed the reference contamination and (3) suppressed the noise recurrence. This has allowed me to produce a 'cloud catalog' that is > 99% complete (as opposed to previous cloud-based flux-recovery of 50-60%). Large-scale summation/averaging of spectra is now possible, with little or no sign of the baseline ripple. The rms noise level in the spectra has been reduced by a predictable amount of ~20-25%.
The first-pass reprocessing was applied to ~60% of the FCRAO OGS (and included most of the brightest emission). An image of a part of the cloud catalog, projected over all velocities (and lacking the cheesegrater) can be seen : HERE. There's some more of it HERE.
An example of the baseline improvement can be seen below.
|Average spectrum over 20X7 degrees. Original dotted, new solid.||Zoom in, to show the previously swamped far outer galaxy clouds at around -80 km/s. And yes... that bump at around -95 km/s (and beyond) is real.|
In the near(ish) future a complete reprocessing will be done. Both the reprocessed (regridded?) data and the cloud catalog will be made immediately available to the CGPS consortium. In the meantime, the FCRAO OGS (barring any low-level reference contamination) is still a sensitive data set, and unless you're doing something large-scale and/or statistical you should have few or no problems. The reprocessed data will not show you something that you couldn't see before (except in cases like in the baseline link given above).
We have used a multiwavelength data set from the CGPS to study the Galactic HII region KR 140, both on the scale of the nebula itself and in the context of the star forming activity in the nearby W3/W4/W5 complex of molecular clouds and HII regions. From both radio and infrared data we have found a covering factor of about 0.5 for KR 140 and we interpret the nebula as a bowl-shaped region viewed close to face on. Extinction measurements place the region on the near side of its parent molecular cloud. The nebula is kept ionized by one O8.5 V(e) star, VES 735, which is less than a few million years old. CO data show that VES 735 has disrupted much of the original molecular cloud for which the estimated mass and density are about 5000 Msun and 100 cm-3, respectively. KR 140 is isolated from the nearest star forming activity, in W3. Our data suggest that KR 140 is an example of spontaneous formation of, unusually, a high mass star.
New radio and optical observations of Galactic surroundings near (l,b)= 94°,2° are presented, revealing new information about the inter- stellar medium and objects in this locale. Of eleven Galactic objects in this area, for example, only two have distance estimates. In this paper, we present new CGPS radio continuum observations of the HII region NRAO 655 (G93.4 + 1.8) at 21cm and 74cm, and optical and radio emission line observations at 656nm and 21cm. The radio spectrum of this object confirms its emission as thermal in origin. From the CGPS HI data we find an atomic hydrogen cavity associated with this object at v = -71.5 km /s. This HI cavity corresponds in position and size to the brightest radio continuum emission from NRAO 655. The corresponding kinematic distance is 8.8 kpc, and NRAO 655's linear size is therefore 70 pc X 130 pc. To confirm the 21cm HI velocity we present the first recombination line detection of NRAO 655 (H158 (alpha) line, v = -72 km/s, width 40 km/s), and the first observations of a molecular cloud interacting with NRAO 655 (at -72 km/s). The first ever optical detection of H(alpha) emission line features is also presented, and the H(alpha) emission line intensity is determined. We find good correlation between optical and radio morphology and between 60 micron infrared and radio morphology. We find anticorrelation between optical and infrared features of NRAO 655. A physical model for NRAO 655 and its environment is proposed.
As part of a systematic study of stellar wind phenomena around massive stars, we (Magdalen, Nicole, myself and our Argentinian colleagues Silvina Cichowolski, Marcelo Arnal and Cristina Cappa) have analyzed CGPS data around the Of star HD 229196, located in Cygnus near the SNR G78.2+2.1. The radio and infrared continuum data show the star to be located in a bay of emission suggesting that the interstellar medium immediately surrounding the star has been shaped by the ionizing flux and stellar wind of the star (to see the radio and IR images, look under Stellar Wind Phenomena on the consortium web page, then under Search for wind bubbles around isolated O stars and click on "AR17"). [Editor's note: This last web address will be changing soon so best not to "bookmark" it]
More interesting however is the presence, in the HIRES infrared data, to the west of the star, of an extended (greater than 30 arcmin in diameter) and highly symmetrical structure which we interpret as a stellar wind bubble. There is no evident corresponding large scale structure in the radio continuum data, however interpretation is complicated by the overlap of the IR bubble with the SNR G78.2+2.1. On morphological grounds, CGPS HI data suggest a possible association with neutral hydrogen at an LSR velocity of about 20 km/s. Although a compact source is clearly visible near the geometrical center of the shell, both in the IR and in the radio continuum at 4.85 GHz (Gregory and Condon 1991), it is however not detected in the CGPS 1420 MHz continuum data and has no obvious optical counterpart.
GSH 138-01-094 is a very conspicuous almost perfectly circular, expanding HI shell in the MX1 field. The radius, expansion velocity, centre and central velocity of the shell were determined by fitting the velocity of the shell material with the expansion velocity-field of a thin shell. The expansion velocity is 11.8 ± 0.9 km/s, and the radius is 37.3 ± 2.2 arcminutes.
The central velocity of -94.2 ± 0.5 km/s suggests a kinematic distance of 16.6 kpc (R0=8.5 kpc, V0=220 km/s). Arguments that the kinematic distance can be used, include the circular shape of the shell, which would be flattened if its velocity through the ISM would be much higher than its expansion velocity. The central velocity is not consistent with streaming motions in the Perseus arm. The large kinematic distance implies that the radius of the shell is 180 parsec, that its mass is 2x105 solar mass and its dynamic age, R/(3.5 V), is 4.3 Myr.
On the low-longitude side, emission of the shell is entangled with an HI cloud which may be related to the shell. A distinct elliptical hole in the shell is found on this side. The orientation and axial ratio of this hole are suggestive of a circular hole, possibly a secondary shell, projected on the plane of the sky.
There is little evidence of the shell at other wavelengths. A previously known small molecular cloud (VLSR=-103 km/s) can be associated with the approaching side of the shell because a chance alignment with such a rare high-velocity CO feature would be unlikely. We have found no evidence of the shell in the 74-cm continuum and 100um infrared images.
The origin of this HI shell is as yet difficult to constrain, and subject to ongoing work. We find that the properties of the shell are in remarkable agreement with the latest stage of the evolution of 1-D hydrodynamic simulations of supernova shells in a low-density ISM with a radiative cooling function for subsolar metallicity by Thornton et al. (1998). This would suggest a very old supernova remnant in the pressure-driven snowplow phase.
Larger renditions (jpegs and tifs) of these icons of the expanding shell GSH138-01-094 can be found on the on the Consortium website at http://www.ras.ucalgary.ca/CGPS/gallery/press/shell/.
The left image includes radio continuum point sources, dust emission, and CO emission as well as HI emission from -93.8 through -110.29 km/s. The right image is pure HI; red is used for a fiducial velocity range and blue has been added to blueshifted velocities (generating pinker colours) while yellow has been added to redshifed velocities (generating oranger colours). An image table on the Consortium website describes the colour assignments in detail.
These images are not yet public. Please abide by the embargo until Stil and Irwin publish their paper. These colour images were constructed by Jayanne English with support from Russ Taylor.
The construction of unofficial CGPS HI "super-mosaics", made by combining the 5°x5° mosaics into larger ones, has shown that the Galactic HI ISM is full of HI self-absorption structures over a large range of angular scales from the resolution limit at 1' up to the degree scale. The situation on several-degree scales is less clear: are there absorption features that large?
The identification of small angular scale HI self-absorption ("HISA") features is relatively straightforward since the bright HI background changes relatively slowly on the angular scale of the absorbing object. The situation is much more difficult for hypothetical many-degrees scale HISA objects because the background varies strongly in position and velocity on those angular scales. It is thus very difficult to pick out such HISA structures.
Therefore, you can imagine our astonishment when, in the first HI super-mosaic we made (the so-called VWXY super-mosaic covering longitude 130° to 147°), we immediately saw an enormous curved "dark arc", one degree wide and ~ 15° in length, stretching across the super-mosaic. The LSR radial velocity in which it appears, ~ -80 km/s, would naively imply that it was located far in the Outer Galaxy some 10 kpc from the Sun. If it were a cold absorbing HI cloud at its kinematic distance, it would be ~ 3 kpc long, ~ 200 pc wide, and contain (in a very crude estimate), ~ 750 000 solar masses. Our analysis suggests that the arc, as an absorption feature, has an HI optical depth of ~ 1 and an excitation temperature of ~ 30 K.
Faced with a strange object with such characteristics, our next, disquieting thought, was that it was some bizarre intrumental or data processing artifact. However, we quickly discovered that the arc was visible in other data sets, including the Weaver and Williams survey, the Leiden-Dwingeloo survey, and in a map of the region from the Effelsberg telescope. The dark arc is thus a real feature.
Two major questions then come to the forefront: was the dark arc an absorbing structure or simply a void bereft of gas? And how far away is it?
We believe that the object is indeed a cold arc of gas rather than a void. Naively, it is easier to picture a cold HI filament or section of a shell rather than a long narrow void of gas, particularly since at its high radial velocity this void would be both ~ 3 kpc in length and have a rather peculiar radial velocity structure: this "void" does not seem to share in the general radial velocity versus Galactic Longitude behaviour of HI clouds ~ 10 kpc away from us. Instead, it appears to be almost stationary in position with radial velocity. It would be difficult to conceive of such a large void of gas in the Outer Galaxy with such a peculiar geometry and velocity behaviour.
If, as we assume from now on, that the arc is a cold cloud, its radial velocity does not necessarily indicate its distance from us. If it had a large peculiar velocity (several tens of km/s) with respect to the Galactic rotation as a whole, it could be much closer (and hence less large and massive) than the naive kinematic distance (well, it could be further, too...). The Perseus Arm contains a number of radio continuum sources, some of which lie on or very near the arc. At present we are unable to detect any absorption towards those sources (from CGPS or the literature) at the arc's radial velocity, suggesting that the arc is located beyond the Perseus Arm. For the present therefore, we simply assume it lies near its kinematic distance.
This object shares many of the characteristics of Carl Heiles' "supershells", whose characteristics and nature are still a puzzle some 20 years since their discovery. The difference here is that a "supershell" candidate is being seen in absorption rather than emission, which permits us a more detailed investigation of its physical properties.
The dark arc is not in Heiles' supershell catalogue, but remarkably, it sits just above one: the HI emission supershell known as GS 139-03-69, which lies below the Galactic Plane, occupies the same longitude range as the dark arc. More remarkable still, GS 139-03-69 (radial velocities ~ -70 km/s) appears to be roughly the same size, shape, and has a similar radius of curvature as the dark arc: in fact they appear to form a single elongated ring or shell. A very simple model might be that the two form a single ring of gas which is expanding and/or is being stretched by differential rotation. The southern part of the ring (GS 139-03-69) is in emission against an empty background at ~ -70 km/s, while the northern part (the dark arc) is in absorption against the Plane at ~ -80 km/s.
We conclude that both objects together form a ring-like supershell in the Outer Galaxy.
An examination of the database of HI spectra taken with the DRAO 26-m Telescope to provide short-spacing data for the CGPS has revealed the presence of a large, double-lobed emission feature at positive velocities near the Galactic plane. Both lobes are at the same "forbidden by Galactic rotation" velocity. The object may be Galactic and represent some energetic jet phenomenon, or may be a pair of high-velocity HI clouds. Further observations are needed to determine the nature of this enigmatic object or objects.
This talk at the CGPS Science Workshop in Montreal set out to demystify polarimetry for the non-specialist, and to illustrate the results obtainable through the use of polarimetry as a tool for the exploration of the ISM. The talk drew on the work of the entire CGPS polarization group at DRAO and the University of Calgary. Much credit goes to Joanne Brown, Peter Dewdney, Andrew Gray, Marta Peracaula, Russ Taylor, and Buelent Uyaniker.
Supernova remnants (SNRs) are polarized emitters, and a few SNRs at high latitudes have been mapped with the DRAO Synthesis Telescope (e.g. DA530, Landecker et al 1999; Cygnus Loop, Leahy et al 1997). However, polarization images from the CGPS closer to the Galactic plane show that SNRs are not prominent. Rather, the dominant features in the images are widespread polarization structures which show no corresponding variations in the total power of the emission. Because the effect is seen predominantly in polarization angle, it is interpreted as arising from irregular Faraday rotation in an intervening magneto-ionic medium (MIM), ionized gas threaded by magnetic fields. We refer to this as the Faraday screen.
The polarized emission on which this Faraday screen operates is the Galactic synchrotron "background" emission, linearly polarized at the point of origin. SNRs are isolated features which must contribute polarized emission, but the polarized emission in the plane covers a much greater fraction of the images than is filled by the SNRs. At 1420 MHz, the state of the polarization when the signal arrives at the telescope tells more about the MIM through which the signal has passed than it tells about the source of the emission itself.
A few general remarks can be made about the Faraday screen. (1) At some longitudes there seems to be more structure at the mid-plane, with the intensity of features declining with increasing latitude, but there are other places in the CGPS area where no latitude dependence is obvious. (2) There are large bar-like features seen in some regions with extents of several degrees. Their origin is not understood. (3) Polarization images made with the Westerbork telescope at 325 MHz show Faraday-screen structure at high latitudes. The DRAO telescope sees little structure there. Conversely, the Westerbork telescope sees little structure in the plane. This results from the wavelength-squared variation of Faraday rotation. Features at high latitude which give significant rotation at 325 MHz produce rotation angles a factor of 20 less at 1420 MHz, below the limit of detectability. Conversely, features in the plane which produce a measurable rotation at 1420 MHz, will produce enough Faraday rotation at 325 MHz to be depolarized (see below for an explanation of this effect). The two telescopes are sensitive to separate regimes of the magneto-ionic medium.
Faraday rotation at a given wavelength is proportional to the line-of-sight component of the magnetic field, the electron density, and the path length. If two of these three quantities can be independently estimated, or at least constrained, electron density or magnetic field may be determined. Magnetic field is of particular interest, because it is hard to measure with other techniques.
An extended envelope of the HII region W4 can be detected by its Faraday rotation (Gray et al, 1999). Because of the high sensitivity of the polarimeter to Faraday rotation, this envelope can be seen by this technique to a greater distance than with other methods. A magnetic field of 20 microGauss was measured in this envelope, corresponding to a field of about 3 microGauss in the surrounding medium if field follows roughly the square root of electron density.
Faraday rotation cannot destroy polarization, but there are a number of effects, loosely described as "depolarization", which reduce the apparent degree of polarization. Generally depolarization is classified under three headings:
Depolarization by HII regions can give information on the distance of the Faraday screen. Using this technique, we have clearly demonstrated that, in some regions, the Faraday screen lies in the Perseus Arm. Correspondence with other structures can be very helpful in locating the Faraday screen. For example, there is a convincing correspondence between the "mushroom" HI structure (English et al 2000) and scale size of polarization features. There appears to be a transition from chaotic field to smooth field which corresponds to the transition from the stem of the mushroom to the cap. A possible role in the formation of the mushroom is not understood.
Other than this example, correspondence between polarization images and HI images has not been convincingly demonstrated, and much more work needs to be done in this area.
Strong depolarization by a foreground HII region effectively eliminates any polarized emission which arises behind the HII region, so that the entire polarized signal must be generated, and Faraday rotated, along the line of sight to the HII region, often at a known distance. This can provide a useful probe of the local ISM. Using this reasoning, Gray et al (1998) have shown that a very regular feature seen superimposed on the HII region W5 is probably caused by an ionized cloud in the interarm region between the Sun and the Perseus arm. The size of the polarized feature is about 2 degrees; its smooth structure stands in marked contrast to the small-scale structure seen in its surroundings.
The telescope makes 1420 MHz continuum and polarization images in four bands, each 7.5 MHz wide. The individual polarizatiom images can be used to make maps of rotation measure. Marta Peracaula is studying the spatial structures in these images.
Rotation measure can also be determined for point sources. Since the sources are predominantly extragalactic, this probes the magnetic field along the whole line of sight through the Galaxy. This work is the subject of Joanne Brown's Ph.D. thesis at the University of Calgary. The high angular resolution and sensitivity of the Synthesis Telescope allow more sources per square degree to be measured (by a factor of 5) than in previous work of this type. Furthermore, previous work has avoided regions close to the Galactic plane.
A quick review of recent HI self-absorption results was given in Montreal. Most of these are outlined in the March newsletter. For more thorough coverage of HISA issues, including a preprint of the July ApJ paper, please visit the HISA page.
By now, Gibson & Taylor had hoped to have high-resolution HI images from WSRT processed, but this has been delayed by a number of unforeseen technical problems with the observations, solutions to which are still being investigated. However, Brunt & Gibson have obtained new FCRAO 13CO data to complement the 12CO data taken in February. This should help in estimating molecular hydrogen column densities toward HISA features.
The figure at right shows detected 12CO emission with 13CO contours -- both integrated over Perseus arm velocities. 13CO was detected in essentially all sightlines with bright 12CO emission. A comparison of 12CO with HISA and H2 implied by dust emission is given in the March newsletter.
Observations of the gas and dust content of high-latitude gas clouds have shown interesting relationships between the emission by the different constituents of these structures. Several studies comparing IRAS, HI and CO emission in these cirrus clouds have yielded evidence of a molecular component which is not traced by the emission of CO molecules (Reach et al. 1994, Meyerdierks & Heithausen 1996, Boulanger et al. 1998). Infrared excess emission from IRAS data points to the possibilty of molecular gas component that is warmer and more diffuse than the component of molecular gas that is traceable by CO. In these studies, the "diffuse" component of molecular hydrogen is found to be comparable in mass to the HI content. This diffuse gas may be widely abundant in the plane of the Galaxy. If present, on the same scales as in these high latitude clouds, it has important implications for many areas of astrophysical interest.
The Canadian Galactic Plane Survey (CGPS) offers an excellent opportunity to study the phenomenon of diffuse molecular gas in the plane of the Galaxy by combined analysis of pc-scale resolution images of CO, HI and dust emission over a large area of the Galactic disk. I will make use of the datasets available to the CGPS consortium to probe interstellar Galactic clouds for molecular gas not traceable by CO emission. The detection of diffuse molecular gas, an analysis of the environmental condition required for its existence, and the correlation with molecular gas traced by CO emission will provide information on the importance of this ISM component and its relationship to the other states of the ISM. The presence of hot stars in the Galactic plane while complicating the study some regions, also yields a warmer and more varied interstellar medium which may give rise to an increase in the amount of molecular gas not seen in CO.
It is now clear that disk-halo interaction is an important aspect of any star forming disk galaxy. Disk-halo features have been observed, for example, in every ISM tracer, including synchrotron emitting particles, molecular, atomic and warm ionized gas, hot x-ray emitting gas, and dust. In ultra-luminous IR galaxies, outflows have been implicated as a primary source of IGM gas and possibly a contributor to a dusty universe. For lower luminosity galaxies, while the issue of disk-halo circulation or disk-IGM outflows has not yet been resolved, it is at least clear that disruption of the ISM internally can significantly affect metallicity gradients, star formation rates, and the global star formation history and hence evolution of the galaxy.
External edge-on galaxies can provide us with a bird's eye view of disk-halo interactions and it has been argued that we can learn much about the Milky Way by studying such systems. Surprisingly, the most common explanation for disk-halo features, i.e. supernovae and stellar winds, has difficulty accounting for all of the observations. For example, high latitude radio continuum emission is found to be present even in galaxies with low star formation rates (Irwin, English & Sorathia 1999). Numerous HI studies have also been undertaken on both face-on galaxies (to detect HI holes) and edge-on galaxies (extra-planar arcs and shells). Again, there are numerous examples of HI features which show no evidence for any internal star forming region, even though the respective timescales require that they be visible (e.g. Rhode et al. 1999, Efremov et al. 1998). Multiple supernovae and stellar winds also fall short of the energy requirements for the largest shells. Indeed, if these are the origin of the kpc-scale HI features, then up to 105 supernovae are sometimes implicated (e.g. King & Irwin 1998)! The energy problem is not new. It was recognized as early as 1979 by Carl Heiles for features seen in our own Galaxy (Heiles 1979, 1984). This suggests that some other source of energy, as yet unknown, must at least be contributing to the larger features.
The alternative explanation for disk-halo features is the impact of high velocity clouds (HVCs). While this alleviates the energy problem, it does not explain the fact that some galaxies in which large HI shells are seen appear to be isolated and that no HI clouds massive enough to produce these features appear to be in the vicinity, even though the observations are sensitive enough to detect them. It is also somewhat disturbing that, although there appear to be no differences in the characteristics of the large shells in comparison to the smaller ones, yet two very different mechanisms must be invoked to explain them.
Theoretical models have been developed for both of these origins and it does appear that generic features are predicted for each case. For example, upright mushroom-shaped features are expected for outflows resulting from internal energy sources such as supernovae (cf. Tomisaka 1998) whereas inverted mushroom-shaped features would be more likely in the event of an impacting cloud (Santillan et al. 1999). (A caveat is that the magnetic field can play a role in either case, possibly significantly altering the shape of the observed feature.) Of course, another clue would be to simply see identify a young star forming region at the base of the feature or the remnants of HVCs or a stream of HVCs!
What does all this mean for the CGPS Phase II and how can CGPS data help to explore these issues?
Arguably, some of the most interesting results from Phase I of the CGPS have been directly related to disk-halo outflow. I am thinking primarily of the Galactic Chimney (Normandeau, Taylor, & Dewdney 1996) and the Galactic Mushroom (English et al. 2000). Not only have features like these never been seen before in the Galaxy, but they have never been seen before in any galaxy. This underscores one of the most surprising and powerful aspects of the CGPS results: the discovery of completely new kinds of astronomical objects.
While external edge-on galaxies can clearly provide a global impression of disk-halo features, the situation is surprisingly complicated when probed in detail. We now have, for example, images of the galaxy, NGC 5775, in every waveband, representing all ISM components (Lee et al. 2000). It turns out that there are so many disk-halo features that there is a good chance they are blended along any given line of sight. This problem is exacerbated for the radio continuum because of cosmic ray diffusion and is also challenging, given the relatively low spatial resolution available for the study of external galaxies. Thus, studying the actual placement of, say, the ionized gas with respect to the neutral gas, which is important for distinguishing between models, is rather difficult. Obviously having a clear view showing where each of these component are, is extremely important for determining the details of disk-halo outflow. A similar problem arises when trying to pinpoint possible star forming (or other) regions at the base of the features, since the line of sight passes through many kpc of disk. One could overcome some of these problems by looking instead at face-on galaxies, but then it isn't so clear whether the HI feature being observed is indeed extra-planar.
Galactic plane studies have their own problems with distance ambiguities, but since the CGPS has primarily focussed on the outer Galaxy, and since the "normal" HI emission falls off with latitude, it does seem to be possible to isolate specific disk-halo features, follow-up with multi-wavelength overlays, and also search relatively cleanly for underlying star forming regions.
Apart from these improvements, the CGPS also provides a very nice complement to extragalactic work in the size scales over which such features are observed. For example, in external galaxies, one can only select the largest, typically kpc-scale features due to the available resolution. In NGC 5775, for example, our highest resolution observations in which disk-halo features can still be detected correspond to a linear scale of about 600 pc. The Galactic Mushroom (English et al. 2000) spans about 350 pc and required some extra CGPS fields at high latitude to complete the mapping. Thus, an increase of factors of, say, 3 in latitude would bridge this gap and allow the discovery of more features of order 500 pc (depending on distance). To look for the larger, kpc-scale features would likely require much more sky coverage than time permits.
Moreover, the detail with which disk-halo features can be mapped in the Milky Way is unprecedented and will certainly not be matched by any external galaxy survey. The HI observations are critical to defining disk-halo features, determining (as best as possible) their distance and especially working through the energy requirements. However, the complementary polarization and radio continuum observations are also important. The orientation of the magnetic field, for example, plays a significant role in the dynamics of disk-halo outflow and, indeed, is a factor in determining whether the outflow may reach beyond the lower halo. Radio continuum observations, if the smooth component can be detected, will also be important for determining the radio spectral index. It has been shown, for example, that there is much more structure in the spectral index map than in the total intensity continuum maps (Duric et al. 1998). Such maps can provide information on the locations of shocks, also important to modeling.
To sum up, all aspects of disk-halo interaction can potentially be addressed by looking at our own Milky Way, including the origin of these features, the details of the models and eventually, global issues such as consistent outflows or circulation. The question is, what emission will the CGPS, Phase II be able to detect?
The pilot fields of the CGPS-II have now been observed and reduced and include Galactic latitudes up to 30 degrees. After some smoothing and averaging of channels, I find that local HI appears to be present in all fields and that emission can also be seen even in the highest latitude field out to about the velocity of the Perseus arm. Beyond the Perseus arm (more negative velocities), the highest latitude field appears quite empty. The contribution of HI to various channels increases significantly as latitude decreases and this has been described quantitatively in presentations by Knee and Taylor. I would expect that, in terms of the HI, good detections can be obtained over many velocity channels for fields up to about 17 degrees latitude. Note that, at a Perseus arm distance of 2.2 kpc, a 500 pc feature corresponds to 13 degrees. Obviously more distant features will appear smaller, but since the HI distribution in the Galaxy falls off with galactocentric radius, it would be best to weight the observations to detect the closer features of somewhat larger angular size. As for continuum observations, little is seen in the total intensity images at the highest latitude other than point sources (though very interesting polarization structures are seen). For disk-halo science, it is the smooth component which is needed, arguing for the addition of the zero-spacing flux and possibly, again, not going to the highest latitudes sampled in the pilot survey. Thus, from a disk-halo perspective, a Phase II project might be best served by thorough mapping up to, but not exceeding a 17° latitude field.
This article reports upon the preliminary results of DRAO observations made as part of the mini-pilot survey for Phase II of the CGPS, specifically of the observations we have made towards the high Galactic latitude nearby low mass star formation complex in the Cepheus Flare Clouds Complex. The Flare Clouds should not be confused with the Cepheus A-E regions, which are high mass star forming regions closer to the Galactic Plane and further away.
As most CGPS Consortium members should be aware, it has been proposed that part of the Phase II observations be devoted to the Cepheus Flare Complex. The general theme of the science case (which has been discussed in greater detail at the 1999 CGPS Science Meeting in Vancouver and in Calgary at the LRPP "Town Hall Meeting") emphasizes that this would be the first really comprehensive multi-ISM component view of a nearby low mass star formation region, which would include both the atomic and molecular phases of the gas, as well as the dust, and perhaps also the magneto-ionic medium. It would be the first such study of an extended complex which would have multi-wavelength coverage over the entire size scale from that of individual star forming cores (~ 0.1 pc, ~ 1') to that of the entire complex (tens of pc, several degrees).
There are other reasons why such a survey should be part of Phase 2. In Canada we have an active and significant community of astronomers who study the ISM and star formation in both the mm and sub-mm as well as by means of theory. It would be to the great benefit to Phase II of the CGPS if some of that community could be attracted to join the Consortium. It would lead to a wider and deeper level of support for the continued CGPS and help expand the client base of the DRAO facilities.
The goals of the mini-pilot survey were first of all to probe the technical feasibility of observing the Cepheus Flare with the current capabilities of the DRAO Synthesis Telescope, and to make a very preliminary assessment of some of the types of science which could be pursued.
The Cepheus Flare Complex comprises a collection of molecular clouds and cores distributed over an area of size ~ 17°x17°, with individual clouds located 200-500 pc from the Sun. Four DRAO fields were observed, all located at Galactic Longitude 114°, and covering the Galactic Latitude range +9° to +22°. The choice of longitude was a compromise with the desire to also probe high latitude 21 cm continuum polarization features, and thus was not optimum from the point of view of the Cepheus Complex. However, one field, designated "LI", covered the major CO cloud located at (l, b) ~ (114°, +17°). This choice also ensured that both the cloud/core and intercloud/intercore medium of the Complex was observed, and thus potentially permitting an assessment of the different types of phenomena a complete survey would reveal.
The fields were observed in the usual way, and were reduced and short-spacing added by Russ Taylor.
In all four fields, HI emission at the several tens of Kelvin (brightness temperature) was observed at Cepheus radial velocities. The latitude-velocity diagram shows a bifurcation in the velocity of the HI: a more widely distributed and fainter HI component shows a pronounced radial velocity gradient from ~ +10 kms/ at l ~ +9° to ~ -5 km/s at l ~ +22°. An enhancement in this gas at ~ -5 km/s is associated with the cloud at (114°, +17°). In this component we have thus successfully detected both HI gas associated with CO clouds as well as the HI in the intercloud medium.
The brighter HI component at ~ -10 km/s appears to be gas restricted to l < +15°. It forms an extended bright region just south of the CO clouds in the L1251 and L1247 region, but appears to correlate morphologically with the IRAS 100 micron optical depth rather than with the CO intensity. This presumably intercore medium is surprisingly bright in HI.
In addition to HI emission, absorption against background continuum sources is fairly commonly seen here, the most obvious examples being towards the sources 3C 454.1, 4C 77.21, GB3 2157+765, as well as towards a hitherto uncatalogued radio source. Some of these absorptions are quite deep, as much as 80 K below the baseline. Studies of these absorptions will help greatly in pinning down the physical characteristics of the gas (i.e. optical depth and excitation temperature).
In addition to absorption against continuum sources, there is clear evidence of self-absorption in the HI gas. Hints of HI self-absorption are seen towards the L1251 and L1247 molecular clouds, but the most impressive example is found towards the CO cloud at (114°, +17°). In this cloud, a patch of HI absorption lies immediately in front of the CO brightness maximum, and remarkably, appears to follow the CO maximum as it slowly shifts on the sky with radial velocity. This is evidence that the HI absorption is intimately connected to the CO cloud.
A more speculative, but very interesting research possibilty concerns the possible association of 21 cm continuum polarization with these nearby high latitude clouds. A rough processing of the polarization data shows that in the fields below ~ 19°, there are polarized intensity structures with angular sizes typically of ~ 0.5° or less. There is a tantalizing suggestion that some of these polarization structures may be positionally associated with regions in the HI halos surrounding the CO clouds.
In summary, the mini-survey has demonstrated the technical feasibility of the DRAO Synthesis Telescope to collect scientifically useful HI data in the Cepheus Flare Clouds Complex. Strong HI emission and absorption features were easily detected over essentially the entire field of view observed. Discrimination between core and intercore HI emission appears possible, and HI associated with individual cores identified. Of particular interest is the HI self-absorption associated with bright CO emission regions, as well as a possible correlation between polarization features and cloud halos. The Cepheus Flare Complex will undoubtedly be a rich vein of interesting science.
It is a pleasant coincidence that the Westerbork Synthesis Radio Telescope (WSRT) produced images at 327 MHz with very nearly the same angular resolution as the DRAO at 1420 MHz. Although the WSRT images do not include low order spacings, for analysis of compact features a direct comparison is straightforward. To facilitate this comparison, Steve Gibson has regridded the data from the Westerbork Northern Sky Survey (WENSS) to the CGPS mosaic geometry. WENSS data for each of the CGPS mosaics are available at the ftp site in Calgary.
One very useful way to visualize the combined data sets is to use the
kview-rgb program in karma. In the image below I have loaded the 327 MHz
image in the red scale and the CGPS 1420 MHz image in the the blue and
green scales. The intensties are adjusted so that pixels that typical background
radio galaxies (spectral index about -0.7 between
327 MHz and 1420 MHz) appear white. Source with inverted (thermal) spectra such as compact HII regions appear blue. Sources with very steep synchrotron spectra show up a red.
|The MW1 field|
|327 MHz (red), 1420 MHz (blue-green)|
One object at upper left has an extremely red spectrum. A close up of the region around this source at 327 and 1420 MHz is shown in the two images below. The source appears only at 327 MHz. At that frequency it has a peak flux density of about 180 mJy. At 1420 MHz the upper limit is about 1 mJy, implying a spectral index of less than -3.5. The only known objects with this steep a spectral index are millisecond pulsars. The object does not appear in any pulsar catalogues. We are now in the process of securing observations for a pulse search to test this hypothesis. We hope this will be the first of many discoveries of exotic compact radio continuum sources.
|Millisecond pulsar candidate at 327 MHz||Millisecond pulsar candidate at 1420 MHz|
In March 2000 I attended the "Stars, Gas, and Dust in Galaxies: Exploring the links" workshop in La Serena. Very good meeting, an interesting gathering. I gave a 15 minute talk about the W4 superbubble/chimney. I can't say I was overly thrilled with the state of the project and therefore with the presentation, but for what it's worth a copy of the paper that will be in the conference proceedings is available from my web page, just follow the links to "Research" and "Publications".
Figures 1, 2 and 3 show the 21cm continuum Stokes I and linear polarisation images of the high longitude end of the CGPS (MVW super-mosaic). When we compare figure 1 and figure 2 it is striking that distinctive dark regions in the Polarised Intensity map (Fig. 2) coincide with the brighter regions of ionized gas in the Stokes I image (Fig. 1). On the other hand, in the Polarisation Angle image (Fig. 3) we can see spread structures of constant or smoothly changing Polarisation Angle. The spatial scale of these PA structures varies from the image resolution limit (2') to several degrees, in a way that seems to be correlated with the presence of thermal emission. For example, in the left and top part of MVW, where there is relatively little Stokes I emission, the angular scales of PA structures are quite large, while closer to the regions of bright thermal emission, the angular scale of PA variations are significantly smaller.
|Figure 1: Stokes I||Figure 2: PI||Figure 3: PA|
The lack of correlation (and most of times the anti-correlation) between polarised emission and Stokes I, indicates that the Polarisation structures we are seeing are imposed by a Faraday Screen onto the background radiation rather than being actual polarised emission objects (see scheme in Figure 4).
|Figure 4: Faraday Screen|
The change in Polarisation Angle imposed by a magneto-ionic medium is where being ne the electron density, , the magnetic field and L the path length of the screen.
Therefore, the study of the Structure of RM (Rotation Measure) can reveal properties of the magneto-ionic component of the Galactic Plane, such as regions of abrupt changes in the direction of magnetic fields. For that purpose, polarisation images of the individual 7.5 MHz sub-bands centered around 1420 MHz are being produced and, from them, Rotation Measure maps are being created. The aim is to develop a parametric description of the scale sizes of the Rotation Measure structures to determine how the spatial properties of the Faraday Screen are related to general properties of the ISM. The approach I am adopting is to associate to every pixel P a corresponding Structure Function defined by:
SF(DP)=<[RM(P) - RM(P+DP)]2>
where RM(P+DP) is the Rotation Measure of a pixel at a distance DP from pixel P and the angle brackets indicate the median value of the inside squared difference over all the pixels at a distance DP from P.
In order to avoid the pixels whose RM fit is not within the confidence level and the point sources, I am using in the calculation of SF only the pixels with corresponding PI between 0.7mJy/beam and 4mJy/beam (these values have been chosen after studying the distribution of RM versus PI). Figures 5 and 6 show the RM maps of MVW respectively before and after blanking out the non valid pixels.
|Figure 5: RM||Figure 6: RM (thresholded)|
The pixels participating in a calculation of SF(DP) belong to a 3 pixels (~1') wide circular ring of radius DP around each pixel P. In order to be able to calculate SF(DP) for the pixels P that have been blanked out, the differences in equation SF(DP) have been done between radially opposed pixels of the shell instead of using the central pixel P in each subtraction. Figure 7 shows an image of the averaged Structure Function from DP=4' to DP=10'. As it can be seen in the image, the structure of RM gets enhanced: darker areas correspond to higher SF, which indicate rapidly changing RM on short spatial scales. Figure 8 shows SF(DP) for two pixels in a dark area (yellow and purple lines) and for two pixels in a light area (red and black lines). The vertical dashed lines represent the interval of DP represented in figure 7. As can be seen in figure 8, SF(DP) seems to randomize around a certain level for shifts DP > 10'. In the near future I intend to explore the spatial behavior of SF versus DP, correlate it with the total power at different frequencies and explore models for the behavior of SF in a turbulent interstellar medium scenario (e.g. Simonetti and Cordes, 1988 rwsi.conf..134S).
|Figure 7: SF||Figure 8: SF versus DP|
Several studies have demonstrated that the Galaxy has a magnetic field with an ordered large-scale structure. We do not, however, know how that field is generated, how it is evolving, or what its overall structure is. It is generally believed that the magnetic field follows the basic pattern of the spiral arms of the Galaxy, yet the field's strength and direction between and within the arms is unknown. Different theoretical models of the generation and evolution of the magnetic field predict different directions. Studying the magnetic field and looking for ``reversals'', where the field's direction changes by 180 degrees, will help identify which model is correct. The focus of my research is to use Rotation Measures from extragalactic point sources in the Canadian Galactic Plane Survey (CGPS), in conjunction with established data from Pulsars to infer information about Galaxy's magnetic field. In particular, I hope to resolve the debate over how many (if any) reversals are present in the outer region of the Galaxy.
The CGPS is able to obtain rotation measures from point sources with a distribution of 1 source for every 1-2 square degrees. My initial results, using 2 mosaics from the high longitude end of the survey and 1 mosaic from the low longitude end, suggest that there are no reversals beyond the solar circle. This disagrees with other previous studies. It will be interesting to see what emerges as more data become available!
1Saskatchewan Naval Observatory
* We have a NASA/CSA proposal under consideration for an ``IRAS 2'' mission to re-survey the infrared sky. In addition, we are exploring the possibility of Y. Cao and C. Kerton carrying out HIRES processing of the IRAS 2 dataset, and C. Brunt re-observing the FCRAO survey, each in exchange for a second Ph.D. They have not yet returned our calls.
We gratefully acknowledge the many insightful comments provided by Michael Oscar2. This work has been involuntarily supported by grants from the Canadian government.
2Director, Surfing Technology Science Institute, Goose Bay, Labrador