Douglas Currie, Kenneth Kissel, , Ed Shaya, Petras Avizonis and Dan Dowling
Astro-Metrology Group, Department of Physics, University of Maryland
College Park, MD 20742

European Southern Observatory, Karl-Schwarzschild Str 2
D-85748 Garching bei Muenchen

A note on Adaptive Optics at ESO

New details of the star formation process have been revealed by a coordinated use of the ADONIS system on the 3.6 meter telescope at La Silla, in combination with data from the WFPC2 Camera of the Hubble Space Telescope. In this very preliminary report, we illustrate some of the unique capabilities of the ADONIS system for high-resolution observations of the stellar formation processes in the near infrared region.

The observations address M16 (a.k.a. NGC 6611, and the Eagle Nebula). This is a molecular cloud in which a small cluster of high mass stars has recently formed. The very high ultraviolet flux emitted by these early-type stars has dispersed or "photo-eroded" surrounding regions of the molecular cloud. Variations in the density of the gas and dust of the cloud have resulted in an uneven irregular erosion, forming the "elephant trunks" seen in the ground-based image shown in Figure 1. Investigation of the details of the resulting structure gives a measure of the resistance of the cloud to the photo-erosion process, which, in turn, is a measure of the density of the various regions of the cloud. While other properties, such as the magnetic field and local temperature may also affect the rate of erosion, this may be one of the most direct methods of measuring the density profile about pre-protostellar objects.

This data was taken under ESO Proposal 57.C-0796 by Douglas Currie with Kenneth Kissell of the University of Maryland and Domenico Bonaccini of ESO. This preliminary image processing has been conducted by Ed Shaya, Petras Avizonis, and Dan Dowling of the UMd. The analysis of this data will be conducted by this group, in collaboration with other members on the WPFC2 IDT team (i.e., Jeff Hester, Paul Scowen and others) and other individuals at the University of Maryland. The observations described here were conducted in early May, 1996 on the 3.6 meter telescope at La Silla using the ADONIS adaptive optics system and the SHARP II NICMOS Camera.

Figure 1 - M16 Nebula. A ground-based image of the nebula obtained by David Malin, illustrates the context of our observations of M16. Observationally, M16 consists of a cluster of early-type, very luminous, very massive young stars and an HII region containing "elephant trunks" or "columns". The early type stars have photo-eroded most of the molecular cloud leaving behind the structures in the HII region. Fluctuations in density in the original molecular cloud caused irregularities in the photo-erosion process. In particular, on the large scale, the "elephant trunks" pointing northwest, i.e., toward the hot stars, are the result of large scale density fluctuations which have shielded portions of the molecular cloud from photo-erosion. On the smaller scale, which is our area of interest, we will see "bumps" and "pimples" caused by the smaller density variations. We believe that this is due to the excess material in self-gravitating pre-protostellar regions.


The scientific objectives of these observations is to provide the observational data for the understanding of tar formation processes. In particular, the three areas of interest to our group at the University of Maryland which defines the objectives of these observations, and will be the target of the succeeding analysis consist of:

Density Profiles of Dust and Gas as a Function of the Stage of Stellar Formation

The dimensions of the features revealed in the Hubble data yield information related to the density profile of the in-falling dust and gas. For example, we can obtain a "characteristic size" related to the current stage of the formation process for a given object. The ESO data allows one to place the object within the normal classification schemes for star formation, i.e., from the classes described by Hillenbrand to the Class described, for example, by Andre (1995).

Direct Evidence of Pre-Main Sequence Objects for the Fainter Components of M16

M16 has long been in interesting region for the search for "pre-main sequence" objects. This work has been developed initially by Walker, et al. (1988), by Chini, et al. (1990), and by Hillenbrand, et al. (1993). In each case, the primary limitations have been resolution, (with overlap of objects due to the extreme crowding), and limiting magnitude in various bands. The combination of the increased resolution and the deeper exposures of the ADONIS data and the capabilities of Hubble would appear to permit a significant extension of the analysis of Hillenbrand.

Investigation of the Very Red source 367 B

Near Walker Star 367, is located a very strong infrared source which does not appear in our V filter data. Because far infrared data is needed in order to distinguish the early stage of the individual star formation regions, the program at the University of Maryland is currently using data which we have obtained using Hubble WFPC2 data, 350 micron data from CalTech Submillimeter Observatory using the GSFC bolometer array, and the millimeter interferometric observations obtained with the Berkeley-Illinois-Maryland (BIMA) array in addition to the data from the observations at ESO.

Figure 2 illustrates the regions observed at a Silla in the context of our recent WPFC2 observations, and Figure 3 shows the context with respect to the existing near infrared observations. Figure 2 shows only the portion which directly corresponds to our Hubble image. This image is a color composite using the J and K magnitudes as the blue and red images. The white "ghost" image is the same region in the ilter for H` emission from our WFPC2 observations. Thus the white "cores" of the stars come from the Hubble data. The regions which were observed at La Silla in J-, H-, and K-Bands are indicated by the outlines.

Figure 3 illustrates the general context of the Hubble Space Telescope observations, as illustrated by recent ground-based observations in H- and K-bands (Hillenbrand, et al., 1993). The published version of the Hillenbrand data covers 225 square minutes to a limiting magnitude of about 14.



Figure 2 - WFPC2 Emission-Line Image. It is our emission-line image of the central portion of the HII region of M16. This was obtained on the WFPC2 of the Hubble Space Telescope on 22 April 1995 (Hester, et al., 1996) and is a color composite in which the blue image is in the H` filter, the green image is ionized sulphur (SII) and the blue image is doubly ionized oxygen (OIII). This color sequence is in order of the energy of ionization. The solid outlines illustrate the boundary of the regions mapped using the ADONIS System on the 3.6-meter telescope at La Silla. The upper region (denoted the "TIP") explores the top of Column III and the region above the column. The lower region, denoted "367", is located near Walker Star #367. The grey circle indicates the outer limit of the ADONIS FoV due to the optical mount for a beam splitter in the ADONIS Optical System.



Figure 3 - Infrared Structure of the WFPC2 Area


Figure 4 - ADONIS Observation of Region 367:



Figure 4 and 5 show the WFPC observations at the bottom and the ESO data at the top. At present, the ESO data, with its infrared sensitivity, shows far more objects than are visible in the Hubble data. Figure 4 illustrates the region near Walker Star #367, the primary target region of interest for our M16 observations at La Silla. The images in the color composite consist of the red for J, green for H, and blue for K-band observations. The white background image (the "ghost" image) is the H’ data from our WFPC2 observations. This region has a large number of bumps where the photo-erosion process is slowed by positive density fluctuations of different sizes. These specific areas are discussed later. The ADONIS image is 37c by 37c which represents four pointings of the 25.6 x 25.6 arc-second FoV of the SHARP II Camera. The resolution (before image deconvolution) is about 280 mas FWHM for the K-band. This is within a factor of two of that which can be achieved in the visible on the Hubble Space Telescope. The red region at the edge of the image is the annular mount which holds the beam-splitter. The red color illustrates the low (300 K) temperature of the emission of this room-temperature object. The high-resolution information obtained by the dual use of Hubble and ADONIS allows a unique probe into the star forming region. Multiple systems are now resolved.



Figure 5 - ADONIS Observations of TIP: Figure 5 illustrates the region around the "TIP" of Column III. As in Figure 4, this figure is composed of both WFPC2 and ADONIS observations. The region contains a number of protrusions of different sizes and very high activity due to the proximity to the blue stars.

Many of the objects detected in the ADONIS are above and outside of the molecular cloud of the tip, indicating that they are not directly associated with the column.

The outline of this image is 60c by 32c which consists of the data from six individual pointings in K-band and four pointings in H-band and J-band. The resolution is about 280 mas (FWHM). This image has the same color combinations as described for Figure 4.

Finally, since the ESO data has a resolution of a few hundred AU, it is critical to obtain the best resolution in order to understand the extended nature of the objects. To this end, we are using the Lucy algorithm for the deconvolution. This can remove the extended, one arc second skirt of the Adaptive Optics PSF due to the correction residuals, as illustrated in Figure 6. The presence in adaptive optics corrected images of an extended halo around the central high resolution object is typical. This has been observed and its cause well understood. It is due to uncorrected higher spatial frequencies, and in some cases to Deformable Mirror actuators print-through, whose regular pattern makes secondary peaks in the PSF. Astronomers who are looking for faint structures right underneath this skirt, will have to familiarize with deconvolution of their data, as well as take care that enough SNR is present in the observed object and structure.

Figure 6 - Image Deconvolution in K-band: In order to utilize fully the higher frequency components available in the ADONIS data, we need to perform an image deconvolution and/or image reconstruction. For the present data, we use our Lucy-Richardson procedures developed for the Hubble Space Telescope modified for adaptive optics images (Currie, et al.1995). This will allow an improved understanding and identification of the extended and double sources. Figure 6A is the image of a (nominally) point source in our K-band data set. The small patch is 3c and the FWHM of the image from ADONIS is 280 mas. This same image is shown in a three-dimensional surface representation in Figure 6C and 6E. The deconvolved images are shown in Figures 6B, D and F. In this case, the FWHM of the deconvolved image is 130 mas. The C-D image pair is normalized to illustrate the improvement obtained in the Strehl ratio (about a factor of 12, one third of the gain coming from the reduction of the width of the core and the remainder coming from the redistribution of the energy in the extended base of the image) The E-F image pair is normalized to illustrate the narrowing of the peak and the extended structure at the base of the ADONIS image. One sees that there are specific features associated with the ADONIS System. The wings out to about 1" are normal for an adaptive optics system. The sub-peaks are probably due to aliasing caused by the regular pattern of actuators in the deformable mirror. This same feature can be seen in the StarFire Optical Range data described in an earlier publication (Currie, et al., 1995). To date we have performed a spatially independent deconvolution but expect to evaluate the spatially dependent effects and then use our algorithm.





1. Andre, P. Low-Mass Protostars and Protostellar Stages. Astrophysics and Space Science, vol. 24, pg. 29, 1995.

2. Chini, R. and Wargau, W. F. Abnormal Extinction and Pre-Main Sequence Stars in M16. Astronomy and Astrophysics, vol. 227, pg. 213, 1990.

3. Hillenbrand, Lynne A., Massey, Philip, Strom, Stephen E. and Merrill, K. Michael. NGC 6611: A Cluster Caught in the Act. The Astronomical Journal, vol. 106, no. 5, pp. 1906-1946, 1993.

4. Walker, H. J. and Wolstencroft, R. D. Cool Circumstellar Matter around Nearby Main-Sequence Stars. The Astronomical Society of the Pacific, vol. 100, pp. 1509-1521, 1988.

5. Hester, J. Jeff, Scowen, Paul A., Sankrit, Ravi, Lauer, Tod R., Ajhar, Edward A., Baum, William A., Code, Arthur, Currie, Douglas G., Danielson, G. Edward, Ewald, Shawn P., Faber, Sandra M., Grillmair, Carl J., Groth, Edward J., Holtzman, Jon A., Hunter, Deidre A., Kristian, Jerome, Light, Robert M., Lynds, C. Roger, Monet, David G., O'Neil, Earl J. Jr., Shaya, Edward J., Seidelmann, Kenneth P. and Wesphal, James A. "Hubble Space Telescope WFPC2 Imaging of M16: Photoevaporation and Emerging Young Stellar Objects." The Astronomical Journal, Volume 111, No. 6, 1996

6. Currie, Douglas G., Avizonis, Petras V., Dowling, Daniel M., Kissell, Kenneth E., O'Leary, Dianne P., Nagy, James G., and Fugate, Robert Q. "Approaches for Image Processing Supporting Adaptive Optics". 1996. In Proceedings (edited by M. Cullum), Topical Meeting on Adaptive Optics, October 2-6, 1995, Garching bei Munchen, Germany, by the European Southern Observatory, pages 299-304.

Return to Astro-Metrology Home Page...

This Page was created by P Avizonis and updated on October 3, 1996. Please send comments to either Dr. Douglas Currie or myself.