THE UNITED KINGDOM INFRARED TELESCOPE
ANNUAL REPORT
1995 AND 1996

2. Scientific Results during 1995 and 1996

2.2. Selected Scientific Results

2.2.2. IRCAM Detection of Extended ``Dark'' Galaxy Halos

P.A. James (Liverpool John Moores University) and M.M. Casali (Royal Observatory, Edinburgh)

It is commonly known that roughly 90% of the mass of the Universe is composed of ``dark matter'', mass that has not been ``seen'' and is known to be present only due to the effect of its gravitational attraction on other visible matter, such as stars. Although not detected optically, much of this dark matter is known to exist in the halos of galaxies. The astronomical literature is full of discussion of the possible nature of this dark material, and the list of candidates that has been proposed includes some of the more bizarre (and in some cases unconfirmed) members of the astrophysical zoo, e.g. brown dwarfs, black holes, mass-bearing neutrinos, MACHOS, and WIMPS.

We have been undertaking a search for extended halos of near-infrared emission around edge-on disk galaxies, with the aim of discovering direct evidence of the massive halo of ``dark matter'' inferred from dynamical studies. Our initial observations yielded only upper limits for the galaxy NGC 100 (Casali et al. 1995), but the detection of R-band halo light around NGC 5907 (Sackett et al. 1994) motivated the further study of this galaxy and others in the near-infrared.

We observed the edge-on galaxies NGC 5907 and NGC 5714 with IRCAM3 on UKIRT in April 1995 and May 1996 using the standard broad-band J and K filters. We measured the gradient of the surface brightness across the frame -- and not the absolute brightness -- by calculating medians for strips of pixels parallel to the disk of the galaxy, producing a profile of the surface brightness as a function of height above the disk. We find gradients orthogonal to the galaxy plane of 7.410 and 1.010 counts second pixel arcsec for NGC 5907 at J and K respectively. For NGC 5714 the corresponding numbersare 8.210 and 1.410 counts second pixel arcsec. (For IRCAM 1 count corresponds to 6 photoelectrons.)

  
Figure 2: (a) Shows the observed gradient in the J-band for NGC 5907. (b) Same for K-band.

Since the measured gradients are very faint, it is important to be able to exclude the possibility that they might be due to scattered light from the galaxy disks and bulges (which are 95 arcseconds away for NGC 5907), caused by internal scattering within the camera. We tested for this by moving a bright star to positions around the periphery of the array, calculated to simulate the disk and bulge light distributions. Whilst some scattered light was detected, it was too small to account for the halo around NGC 5907, by a factor of 20.

It appears unlikely that the measured gradient is due to the observed disk, a thick disk or an extrapolation of the observed halo. If we identify the detected emission with the component postulated to give rise to the measured flat rotation curve, i.e. a ``dark'' halo, it is possible to estimate an overall mass-to-light (M/L) ratio for the objects comprising this halo. We assume a power-law form and find J- and K-band surface brightnesses at 95 arcseconds (5.2 kpc) of 24.3 and 23.0 mag/arcsec respectively, for NGC 5907. These yield a J-band M/L ratio of 220 and K-band M/L ratio of 100 in solar units, with some dependence on the assumed power law. For NGC 5714, the corresponding values are 170 and 60 at J and K respectively. Our method is insensitive to any constant surface brightness component, and thus an extended central core to the ``dark matter'' halo would not be detected. If a central core exists then the M/L values quoted above would be decreased.

The J-K colour of our detected dark halo can be derived directly from the ratio of the gradients detected in the two bands, and are J-K= 1.300.3 for NGC 5907, and 1.5 0.3 for NGC 5714. Note that this colour is independent of the assumed form of the halo. However, Lequeux et al. (1996) note that the optical colours of the NGC 5907 become redder with radius. This could cause the K gradient, and hence flux, to be underestimated relative to the J values, and the intrinsic J-K colours could thus be even redder than we observe.

By combining our measurements with the R band detection of Sackett et al. (1994) we can determine an optical-infrared colour for the NGC 5907 halo. However, in this case the K surface brightness, and therefore R-K colour, depend on the index assumed for the halo. Taking an index of 1.0, R-K for the halo is 3.5; for an index of 1.2, R-K = 3.3.

At this stage it is not possible to say what types of objects make up the ``dark matter'' halos of galaxies. However, the detection of emission clearly strengthens the possibility that stars and/or brown dwarfs may constitute these halos. If so, the M/L ratios and the near-infrared colours appear to be dominated by stars at the lowest mass limits for hydrogen burning; and the optical-infrared colours require the presence of at least some higher-mass stars in the halo.

References

Casali M. M., James P.A., 1995, MNRAS, 274, 265
Lequeux J., Fort B., Dantel-Fort M., Cuillandre J.-C., Mellier Y., 1996, A&A, 312, L1
Sackett P. D., Morrison H. L., Harding P., Boroson T. A., 1994, Nature, 370, 441