PLANETARY NEBULAE vis--vis GLOBULAR CLUSTERS



Many of the dynamical analyses that we will attempt with plantary nebulae (PNe) can and have been done with globular clusters (GCs), which are also found in large numbers in the outer parts of galaxies. The use of these data can be quite complementary: where there is overlap between PN and GC data, they can be used as cross-checks on each other, and also as reinforcing constraints. However, in many respects, using PNe is more advantageous. The primary reason for this is that the PNe directly trace the properties of the bulk stellar population comprising an early-type galaxy, while the GCs represent a distinctly different population with different properties and formational histories. The table below summarises some points of comparison:


Planetary Nebulae Globular Clusters
The PNe allow us to address directly questions about the normal stellar population of a galaxy, such as its angular momentum at large radii, and orbital anisotropy -- data which provide key inputs into galaxy evolution models. While the properties of the GC systems are interesting in their own right, shedding light on their formation, it is a questionable assumption to take their properties as representative of the stars in general at comparable radii.
PN kinematics at large radii can be tied directly to the stellar absorption line kinematics at small radii, thus providing a single dynamical tracer all through the galaxy. There are typically not enough GCs in the inner parts of a galaxy to probe those regions, and combining the GC with the stellar constraints is not as powerful as if their kinematics were directly coupled.
The average spatial distribution of the PNe is quite well known from the integrated stellar light. The GC spatial distribution is more uncertain, as it must be derived from discrete observations of the GCs themselves. This introduces larger uncertainties in any dynamical models.
Velocity measurements of PNe are very straightforward, as there is simply a single emission line at 5007Å to be measured, resulting in typical measurement uncertainties of ~20 km s-1. This advantage, and the previous two above, mean that a given number of PN velocity measurements will provide stronger dynamical constraints than a comparable number of GC velocity measurements. Velocity measurements of GCs require careful use of stellar templates, resulting in typical measurement uncertainties of 50-100 km s-1.
The amount of telescope time required to attain a sufficent number of velocity measurements (~ 100-200) around a galaxy is roughly the same for PNe or GCs. However, lower-luminosity galaxies are typically more quickly studied with PNe, since the specific number of PNe per unit luminosity increases for fainter galaxies, while it decreases for GCs. Galaxies like M87 and M49 with a large spatial extent are typically more quickly studied with GCs than with PNe using counter-dispersed imaging, since the latter would require taking multiple fields.
PNe can be found and have their velocities measured in a single observation, using counter-dispersed imaging. Starting from scratch, this gives a substantial savings in observing overhead. Many elliptical galaxies already have large numbers of GCs identified through imaging surveys, meaning that the efficiency of follow-up spectroscopy is somewhat more predictable than for PNe, for which relatively fewer imaging surveys have been made.
PNe are observable in adequate numbers to typically ~5 effective radii. Because of their more extended distribution, GCs may be observable to ~10 effective radii, thus better probing the largest radii.




last modified by AJR, 10 October 2002