Filling the Gaps in the GLIMPSE360 Survey

Transcription

1 Spitzer Space Telescope General Observer Proposal # Filling the Gaps in the GLIMPSE360 Survey Principal Investigator: Barbara A. Whitney Institution: Space Science Institute Electronic mail: Technical Contact: Christer Watson, Manchester College Co Investigators: Christer Watson, Manchester College Edward Churchwell, University of Wisconsin Robert Benjamin, University of Wisconsin Whitewater Marilyn Meade, University of Wisconsin Madison Brian Babler, University of Wisconsin Madison Matthew Povich, The Pennsylvania State University Thomas Robitaille, Harvard Smithsonian Center for Astrophysics Science Category: Galactic: galactic structure Observing Modes: IRAC Post Cryo Mapping Hours Requested: 46.4 Proprietary Period(days): 0 Abstract: Regarding the GLIMPSE360 Exploration Science project that is mapping the outer Galaxy, we have some good news and bad news to report. The good news is that the data are proving to be spectacular. Due to the higher dynamic range and the lower confusion compared to the GLIMPSE survey (of the inner Galaxy), we can detect stars out to the edge of the Galaxy, PAH bubbles at 3.6 microns; and YSO outflows, H II regions, and supernova remnants at 4.5 microns; and we can separate IR excess sources in color magnitude space. The bad news is that due to a variety of conditions we did not fully anticipate, a substantial portion of the survey observed so far (about 1/6 of the entire survey) has narrow gaps in coverage, in between AORs. We have corrected the problem and our recent data show no gaps. The complete coverage of this survey is meaningful for several science goals that are based on global studies, such as mapping Galactic structure and measuring star formation rate as a function of Galacto centric radius. We request additional time to fill in these gaps and to realize the full potential of the survey.

2 GLIMPSE360 Gaps, B. Whitney et al. 1 1 Science Plan 1.1 Scientific Justification Overview of GLIMPSE360 GLIMPSE360 is an Exploration Science project that was awarded 1980 hours to map the remaining 183 degrees in longitude of the Galactic plane that have not been mapped by the other Spitzer Galactic Plane surveys (GLIMPSE, GLIMPSE II, GLIMPSE3D, Vela Carina, SMOG, and Cygnus-X). The specific Galactic longitudes covered by GLIMPSE360 are l = 65 76, , and Thus GLIMPSE360 will complete the full circle of the Galactic plane. We will refer to the previous suite of surveys as GLIMPSE, and the current one as GLIMPSE360. The combined Spitzer surveys of our Galactic plane provide a panoramic view of our Galaxy that is both visually and scientifically stunning (see our zoomable web browser showing the GLIMPSE and MIPSGAL surveys at this website: With GLIMPSE (and the MIPSGAL survey) we have calculated the star formation rate of the Galaxy (Robitaille & Whitney 2010), refined the dimensions of the Galactic bar (Benjamin et al. 2005), studied large star formation regions in detail (Povich et al. 2007, 2009, 2010) and cataloged millions of objects, including stars, IR-excess sources, PAH bubbles, outflows from massive young protostars, globular clusters, and external galaxies in the zone of avoidance (see review by Churchwell et al. 2009). Similarly, with GLIMPSE360 we will determine the star formation rate of the outer Galaxy, determine the edge of the Galactic disk, map the Perseus and far outer arm, and look for evidence of star formation in the Far Outer Galaxy. In addition, we and others will catalog many of the same types of objects as found in GLIMPSE. Following the tradition of the previous GLIMPSE Legacy programs, we will deliver enhanced data products (sources lists and cleaned mosaics) to the community. The Legacy value of GLIMPSE and GLIMPSE360 are further enhanced by surveys of the Galactic plane at other wavelengths, such as 2MASS and UKIDDs in the near-ir (Skrutskie et al. 2006, Lucas et al. 2008), WISE at 3-23 µm (at lower resolution than Spitzer), SCUBA-2 in the submm, and the FCRAO CO surveys of the Outer Galaxy (l = , Heyer et al. 1998; l = , Bruno & Heyer, in preparation). The following sections describe gaps in the initial survey observations ( 1.1.2), the need to fill in the gaps ( 1.1.3), and preliminary results from GLIMPSE360 showing that the promise of the survey is exceeding expectations ( 1.1.4) Gaps in the survey We have completed about 1/3 of our survey as of April 7, 2010, covering 66 of longitude (out of 183 ). Due to different observing modes in the warm mission, to Spitzer Science Center s (SSC s) need to have more flexibility in scheduling our observations (less timing constraints), and to larger roll angle changes in the outer Galaxy than we had seen in the inner Galaxy, our initial observing plan produced some gaps in the survey images. These occur mainly in longitude ranges l = , , and The Spitzer roll angle changes at 1.0 degree/day at l=90. Typical roll angle changes in the original GLIMPSE survey region were 0.3 degrees/day and were never as high as is common around l=90. As a result of these quickly changing roll angles, the long strips (used in all the GLIMPSE-related projects) were rotated slightly beyond their tolerance and then gaps be-

3 GLIMPSE360 Gaps, B. Whitney et al. 2 Figure 1: A region centered at (l = 147,b = 0.4 ) showing gaps between AORs in the survey coverage at 3.6 µm. tween the strips resulted. Examples are shown in Figure 1. We did plan 50 extra hours of contingency time (3% of the total) to handle situations like this, but they were used to fix some larger gaps produced by an error in our planning software. Several hours of observing and scheduled observing (that couldn t be changed) occurred before we saw these problems. We have worked with SSC to adjust our overlaps and timing constraints, and our most recent observations contain no gaps. The gaps occurred because we are pushing the limits of efficiency. If we had insured that there would be no gaps by increasing overlap between adjacent long strips, our observing efficiency would have decreased, resulting in a smaller survey area. Based on the success of previous GLIMPSE-style projects, we were aggressive in using the same overlap despite the quicker roll-angle changes. We regret that we did not anticipate these changes, and humbly ask for additional observing time above the original generous allotment that was awarded for this project. The value added is very high as described in the next section. As described in 1.2, the observing plan has no timing constraints The Need to Fill in the Gaps A large value of the GLIMPSE surveys has been the complete coverage over the specified survey area, as the following scientific results illustrate: 1) Counting stars by measuring the slope of logn-logs (source counts vs flux) space allows us to see overdensities of red clump stars. These are excellent standard candles (M4.5=-1.62+/-0.03), and their over-densities are mapping Galactic structure. Benjamin et al. (2005) used this method to map the Long

4 GLIMPSE360 Gaps, B. Whitney et al. 3 Bar of the Galaxy, the three dimensional structure of the inner bar, and over-densities in the stellar disk, possibly associated with spiral structure (Benjamin et al., in prep). 2) Robitaille et al. (2008) produced a highly reliable and complete flux-limited catalog of IR excess sources in GLIMPSE and GLIMPSE II, and identified a likely population of YSOs through color-magnitude selections. These were compared to a population synthesis model of YSOs in the Galaxy to which the same sensitivities, color-magnitude selections, and sky coverage was applied. While we only have a 2-D view of our Galaxy, a population synthesis model that collapses 3-D information to the same 2-D view as the GLIMPSE survey allows us to probe the third dimension in a statistical sense. The variable in the model is the star formation rate, which was calculated to be M /yr (Robitaille & Whitney 2010). Both of these examples make use of a complete and easily specified survey coverage area, sometimes summing over latitude in the analysis. This can be done with gaps in it but would require more effort in the analysis to account for them. For example, the YSO population synthesis model can use a similar mask for survey coverage area as the GLIMPSE360 data. 3) Another example is the set of studies of the M17 star formation region done by Povich et al. (2007, 2009, 2010). These regions cover an area of 2 square degrees on the sky. The M17 molecular cloud is in the process of passing through the Sagittarius spiral arm and is an ideal location to study both local and global triggering of star formation. Povich has shown that the young stellar populations in three different regions form an age sequence across this large complex, as predicted by Elmegreen & Lada (1976). Studies such as this require complete coverage of large areas on the sky. 4) We and others have cataloged globular clusters and external galaxies, outflows from massive young protostars, PAH bubbles from massive young stars, YSOs and AGB stars, and over 100 million stars. The great value of these catalogs is that they are based on complete coverage. We could replan the rest of our survey to make it narrower in latitude, and use the extra time to fill in the gaps. We have already done this when we discovered our earlier, larger gaps. Our survey width is now 2.58, compared to the original 3.1 proposed. The outer Galaxy is thicker than the inner and we are very reluctant to make the survey any narrower (the original GLIMPSE width is 2 from l = 5 65 and up to 8 at the Galactic center). Finally, the images are much more beautiful without gaps in them, and this has intrinsic value to us and to the general public GLIMPSE360 Preliminary Results The GLIMPSE360 observing strategy differs from GLIMPSE in that we observe with only 3.6 and 4.5 IRAC bands, and we observe each portion of the sky 3 times in High-Dynamic-Range (HDR) mode (12 sec and 0.6 sec frametime). The total exposure time on each position in the sky is 13 times longer than the GLIMPSE 2-visit 2-sec frametime exposures. The effective magnitude ranges, based on our recently processed catalogs, are at 3.6 µm, and for 4.5 µm. Note: These are based on single-frame photometry and will go deeper when we do mosaic photometry. We expect those catalogs to be about 2.5 mags deeper and 1 mag brighter than the GLIMPSE survey catalogs. Figure 2 shows updated logn-logs slope plots including the processed portion of the GLIMPSE360 survey. GLIMPSE360 shows preliminary evidence for a slope change coherent in longitude-magnitude space from l = at m = If this is due to red clump stars, it indicates a region of stellar overdensity that is 7-9 kpc distant, consistent with the distance of the Outer Arm. Modelling is in progress to test this interpretation.

5 GLIMPSE360 Gaps, B. Whitney et al. 4 Figure 2: Identifying red clump star overdensities in the Galaxy using 4.5 µm star counts. These show up as changes in the slope of the logn-logs histograms. The apparent magnitude of these inflection points maps to distance. The region highlighted by the yellow question mark maps to a distance of 7-9 kpc. The Perseus arm at 2-4 kpc is not apparent. It will be very interesting to see what more complete coverage will show us. Other features in the GLIMPSE survey are indicated at center. Interestingly, there is no clear feature at the expected distance of the Perseus Arm; further data will shed more light on these surprising results. Figure 3 shows images of star forming regions. These show two important effects of value to our science goals: 1) Protostellar outflows are detected as excess 4.5 µm emission in 2- or 3-color image displays (IRAC 3.6 µm µm, or K + IRAC 3.6 µm µm, respectively). These show up as smallish ( 10 ) green objects in 3-color displays (3.6 µm in blue, 4.5 µm in green, 8.0 µm in red) in the GLIMPSE survey. Cyganowski et al. (2008) has published a catalog of over 300 of these Extended Green Objects (EGOs) and confirmed in followup submm observations that they are outflows from massive YSOs in their earliest stages of formation. This is a rare stage to observe because it is so short-lived. Thus identifying these outflows provides a way to locate these rare objects, and do detailed followup studies to learn more about their physical properties. Using the GLIMPSE360 data, where we do not have 8 µm images, we display the 4.5 µm image in the red channel (Figure 3) so the outflows will have to be renamed to be Extended Red Objects. The image at left in Figure 3 shows one probable outflow (at center in red) and one possible (at the left of the image) from the star forming region WB89 43 (Wouterloot & Brand 1989; also IRAS ). This is located at l = 92.67, b = 3.07, at a kinematic distance of 1.5 kpc. The image at right shows several outflows near the well-studied intermediate-mass protostar GL 490. These outflows are all new discoveries, except one that was discovered in the radio (top right) (Lyder, Belton, & Gower 1998). These also show a second feature of our survey that will enable great science: 2) The higher sensitivity of GLIMPSE360 and the lower backgrounds allow detection of fainter outflows such as these. It also allows detection of PAH bubbles in the 3.6 µm band, as shown in Figure 4, left panel. Churchwell et al. (2006, 2007) identified almost 600 PAH bubbles using the 8 µm images. These are signposts of recent massive star formation and

6 GLIMPSE360 Gaps, B. Whitney et al. 5 Figure 3: Left. The star formation region WB89 43 showing at least one outflow from a massive young protostar, in red at the center. Right: Star formation near GL 490 (which is the saturated source at l = ,b = 1.93 ) showing previously undiscovered outflows (encircled in green). In both images, K (2.2 µm) is displayed in blue, 3.6 µm in green, 4.5 µm in red. good places to study triggering of current star formation (Watson et al. 2008, 2009). In GLIMPSE, the bubbles are not as apparent at 3.6 µm as at 8 µm. In GLIMPSE360, due to the higher sensitivity (more and deeper exposures) and lower Galactic background emission, the PAH bubbles are very apparent at 3.6 µm. Again, this identifies interesting regions that can be studied further. Figure 4 shows the supernova remnant HB3 in red emission in the right panel. Several star clusters are present throughout the images (e.g., Figure 4, left panel, at right in the image). A search for new clusters is in progress. Figure 5 shows that we can separate IR-excess sources from naked stellar atmospheres in color-magnitude space. These IR-excess sources consist primarily of YSOs, evolved stars, and unresolved galaxies. Because of the lower confusion in the outer Galaxy compared to GLIMPSE, the IR-excess sources identified in a given star forming region are more likely to be YSOs and to be associated with that star forming region. In the GLIMPSE survey (of the inner Galaxy), there is more confusion with background and foreground YSOs as well as evolved stars (mainly Asymptotic Giant Branch, or AGB stars). Not shown here, we examined a (GLIMPSE360) region with an overdensity of resolved galaxies. This also showed a large population of IR-excess sources compared to an empty region, likely due to unresolved galaxies in a cluster. The population of IR-excess sources in the empty regions is most likely AGB stars. These are distributed more uniformly in space than YSOs and galaxies, and their source density can be used to study their properties and to estimate their contamination to star forming regions and galaxy clusters. In our YSO population synthesis model, we can make the same color-magnitude selections and sensitivity cuts to our model as the GLIMPSE360 data and compare the results model and data catalogs. We can estimate the galaxy contamination using the SWIRE survey

7 GLIMPSE360 Gaps, B. Whitney et al. 6 Figure 4: Left. PAH bubbles seen at 3.6 µm in green (K in blue, 4.5 µm in red). Right. A supernova remnant seen at 4.5 µm in red (3.6 µm shows PAHs in green, and K is blue). Figure 5: Left. Color-magnitude diagram of an empty field region. The purple points are from the GLIMPSE360 catalog. The greyscale is expected colors of YSOs at the distance of the star forming region GL490 (d = 1 kpc). The black line shows the probable dividing line between naked star colors and those with dusty envelope producing an IR excess (the purple sources to the right are probably galaxies). Right: Color-magnitude diagram of the region surrounding GL 490. The purple points to the right of the black line are probable YSOs.

8 GLIMPSE360 Gaps, B. Whitney et al. 7 and the AGB contamination from source counts in empty regions (after galaxy removal). The main variable of the model will again be the star formation rate. By comparing the longitudinal distribution of the model and data, we ll constrain the Galacto-centric radial distribution of YSOs. As discussed in the original GLIMPSE360 proposal, we have several other science goals that we are confident will be achievable, including finding Infrared Dark Clouds using extinction mapping from SED (Spectral Energy Distribution) fits to catalog stars; searching for supernovae remnants, star clusters, planetary nebulae; and studying the AGB and YSO population in the Outer Galaxy. Since these are global studies, they will benefit from a complete and uniformly sampled survey. We have discussed just a few of the science goals that our particular group is most interested in. We have treated the GLIMPSE360 project similar to the previous GLIMPSE legacy projects in that we are making our processed images and catalogs publicly available. The legacy value of the project will be greatly enhanced by filling in the gaps. 1.2 Technical Plan Observing Plan The regions to be observed consist of thin strips (typically less than 100 in width) and of variable length (between a 1 to 1 ). Our observing strategy makes exclusive use of the fixed cluster-position observing mode. By using positions on the sky offset along a line in increments of 2.4, we are able to assure full coverage (1-3 times) of thin strips regardless of roll angle. This last quality allows our observing strategy to be implemented with efficiency and without any observing constraints of any kind, a marked contrast from previous GLIMPSE-style proposals. Some gaps in the GLIMPSE360 survey resulted from AORs of adjacent segments not overlapping, resulting in a gap that is more square-shaped than the slivers between adjacent AORs within a single segment. These areas are covered in this proposal by AORs designed by hand to be nearly square-shaped and will cover the gap at any roll angle or observing date. We have placed no constraints on these AORs. Our exposure times are the same as for the GLIMPSE360 project (12 second HDR frames) with 1-3 visits on each sky position, depending on roll angle (2-3 visits in most cases) Management Plan The observations have been planned by co-i Watson. Watson will also oversee the datataking. The data will be processed by co-is Meade and Babler. The gap data will be added into the previous data, and the regions will be reprocessed as a set. PI Whitney will manage the project as a whole. The other co-is (Churchwell, Povich, Robitaille) are science users of the data (along with Watson and Whitney). Their heavy use of the data helps to improve the products.

9 GLIMPSE360 Gaps, B. Whitney et al References Benjamin, R. A. et al. 2005, ApJL, 630, L149 Cyganowski, C. et al. 2008, AJ, 136, 2391 Churchwell, E. et al. 2006, ApJ, 649, 759 Churchwell, E. et al. 2007, ApJ, 670, 428 Churchwell, E. et al. 2009, PASP, 121, 213 Elmegreen, B. G. & Lada, C. J. 1976, AJ, 81, 1089 Heyer, M., et al. 1998, ApJS, 115, 241 Lucas, P., et al. 2008, MNRAS, 391, 136 Lyder, D. A., Belton, D. S., & Gower, A. C. 1998, AJ, 116, 840 Povich, M. S. et al. 2007, ApJ, 660, 346 Povich, M. S. et al. 2009, ApJ, 696, 1278 Povich, M. S. & Whitney, B. A. 2010, ApJL, 714, L285 Robitaille, T. P. et al. 2008, AJ, 136, 2413 Robitaille, T. P., & Whitney, B. A ApJL, 710, L11 Skrutskie, M., et al. 2006, AJ, 131, 1163 Watson, C. et al. 2008, ApJ, 681, 1341 Watson, C. et al. 2009, ApJ, 694, 546 Wouterloot, J. G. A., & Brand, J. 1989, A&AS, 80, 149

10 GLIMPSE360 Gaps, B. Whitney et al. 9 2 Brief Team Resume PI B. Whitney, Senior Research Scientist at Space Science Institute, is PI of the GO6 Exploration Science program GLIMPSE360, a member of the GLIMPSE and SAGE legacy teams, and one of the developers of the GLIMPSE data processing pipeline. She developed a multi-dimensional radiation transfer code that has been used to model the SEDs of thousands of YSOs observed by Spitzer. Co-I C. Watson, Associate Professor at Manchester College, member of the GLIMPSE teams and designer of their observing programs. He studies feedback in massive star forming regions. Co-I E. Churchwell, Professor Emeritus at the University of Wisconsin, is PI of the GLIMPSE I and GLIMPSE II legacy projects, and is an expert in massive star formation. Co-I R. Benjamin, Associate Professor at University of Wisconsin-Whitewater, is a member of the original GLIMPSE team, and the PI for GLIMPSE-3D Legacy project. He has led the effort to use GLIMPSE data to constrain the stellar structure of the Galaxy. Co-I M. Meade is a member of the GLIMPSE and SAGE legacy teams, and one of the developers of the Wisconsin IRAC pipeline. She has run the IRAC pipeline on the GLIMPSE and SAGE data, producing source lists and images for the astronomical community. Co-I B. Babler is a member of the GLIMPSE and SAGE legacy teams, and is our expert on photometry, source catalogs and errors. Co-I M. Povich is an NSF Astronomy & Astrophysics Postdoctoral Fellow at the Pennsylvania State University. He has analyzed several massive star forming regions in the GLIMPSE and Vela-Carina surveys using YSO radiation transfer models. Co-I T. Robitaille is a Spitzer Postdoctoral Fellow at Harvard- Smithsonian Center for Astrophysics. He developed a large grid of YSO models and a YSO population synthesis model to calculate the star formation rate of our Galaxy. He produced a GLIMPSE catalog of 20,000 IR-excess sources. Relevant Publications 2010 M. S. Povich & B. A. Whitney, Evidence for Delayed Massive Star Formation in the M17 Proto-OB Association, ApJL, 714, L T. P. Robitaille & B. A. Whitney, The Present-Day Star Formation Rate of the Milky Way Determined from Spitzer-Detected Young Stellar Objects, ApJL, 710, L E. Churchwell, B. L. Babler, M. R. Meade, B. A. Whitney, R. Benjamin et al. The Spitzer/GLIMPSE Surveys: A New View of the Milky Way, PASP, 121, C. Watson et al., IR Dust Bubbles: Probing the Detailed Structure and Young Massive Stellar Populations of Galactic H II Regions, ApJ, 681, T. P. Robitaille, M. R. Meade, B. L. Babler, B. A. Whitney et al. Intrinsically Red Sources observed by Spitzer in the Galactic Mid-Plane, AJ, 136, E. Churchwell, M. S. Povich, D. Allen, M. R. Meade, B. L. Babler, R. Indebetouw, B. A. Whitney et al., The Bubbling Galactic Disk, ApJ, 649, R. A. Benjamin, E. Churchwell, B. L. Babler, R. Indebetouw, M. R. Meade, B. A. Whitney, C. Watson et al. First GLIMPSE Results on the Stellar Structure of the Galaxy, ApJ, 630, L B. A. Whitney, et al., A GLIMPSE of Star Formation in the Giant H II Region RCW 49, ApJS, 154, 315

11 GLIMPSE360 Gaps, B. Whitney et al Summary of Existing Programs PI. B. Whitney is PI of Cycle-3 and Cycle-4 Theory proposals (PID & 40794) to make a large YSO grid and SED fitter publicly available and to modify the codes and produce the next generation grid of models. The tasks for the first proposal are finished, we have published 18 papers, and are currently running the new grid of models. Whitney is also PI of the Cycle-6 GLIMPSE360 Exploration Science program (PID 60020). Data taking started in Sept. 2009, is continuing through 2010, and we are processing the data as it comes in. Co-I E. Churchwell is PI of 4 Spitzer programs. The GLIMPSE I & II programs are completed, with final data products released to the community. 59 papers have been published by the GLIMPSE team, with at least another 143 by the community. For PID 40002, data have been reduced, and a publication is in preparation. Data for proposal 50130, data have been reduced and a publication is in prep. Co-I B. Benjamin is PI of the Cycle-3 GLIMPSE-3D Legacy project (30570), which completed all required data delivery last year. A paper is in preparation. 4 Observation Summary Table Below is a table outlining our observations. Note that we did use the Perl script to generate information from the AOR file but that created a table 4000 lines long. The Spitzer help desk advised us that this smaller table would be sufficient. Longitude range Number of AORs observing time (hrs) The total time to image the gaps is 46.4 hours. The AORs we are submitting are final. 5 Modification of the Proprietary Period We waive the proprietary period. This is an add-on to the GLIMPSE360 project which we view as a legacy project. 6 Summary of Duplicate Observations The GLIMPSE360 project overlaps several large surveys at the boundaries (SMOG, Cygnus- X, and Vela Carina). There are multiple smaller programs that were reobserved by GLIMPSE360, because avoiding them would make the survey very inefficient. These include the ARGUS program (of sufficiently different sensitivity to not be considered a duplicate), W5, and the Galactic First Look Survey. Our gap observations may overlap some of these, as well as overlapping the GLIMPSE360 survey, which is intentional for full coverage.

12 GLIMPSE360 Gaps, B. Whitney et al Summary of Scheduling Constraints/ToOs There are no scheduling constraints on our observations.