If a model is to be considered a successful representation of a
physical system, it must reproduce the essential features of the
system as gauged by observations. In the case of interacting galaxies,
this reduces to matching the characteristic morphology and kinematics,
but it is by no means a simple task.
As outlined in the previous section, there are several parameters
which govern the evolution of an interacting pair of galaxies - the
mass ratio of the pair, the inclination of their relative orbits, the
pericentric argument, the ellipticity of the orbits and the viewing
angle to name but a few. Subtle changes in a single parameter may have
a significant effect on the morphology and kinematics of the observed
pair. Furthermore, it is possible that no single set of initial
conditions uniquely determine the final outcome (Barnes, 1988).
With these significant factors in mind, it is clear that model
matching is a potentially daunting task. However, by making some plausible
assumptions and deciding to work with carefully chosen characteristics
of the system, it is possible to simplify the process somewhat.
When we set out to match tailed galaxies, our sample is
already predefined - spiral galaxies which possess dynamically cold
gas rich disks. As described above, Barnes noted in his 1988 paper
that such dynamically cold disks are essential if the characteristic
features of tidal interactions are to be reproduced. Furthermore, it
is a well known observational fact that spirals have copious amounts
of HI in their disks and that this neutral hydrogen extends out to
approximately twice the luminous radius of a galaxy. This key fact
coupled with the 21cm spectral line of HI which freely exhibits a Doppler
shift in response to its motion offer us an effective
tracer. Such a tracer reveals not only the spatial structure
of bridges and tails but also the line of sight velocities. This
information is vital.
In modeling encounters between galaxies, how the galaxy is represented
can play a important role in how the interaction evolves. As Barnes (1988)
noted, the inclusion of the massive dark halo in his fully N-body self
consistent BDH encounter had a profound effect on the merging time when
compared to Toomre and Toomre's 3 body test particle encounter. (This
demonstrates the importance of the dark halo for tidal friction and
orbit decay.) Any recipe for model matching that is adopted
must target characteristic features which are model independent.
For the model matching, I made extensive use of John's Identikit
display package; this gives 3 separate projections - X-Y, Vz-Y and X-Vz- of the data under consideration, where X,Y and Z are the standard
spatial coordinates and Vz is the line of sight velocity. The user can
perform rotations about the spatial axes as well as scalings (both radial and
velocity) of the model data compared to preloaded HI observational data.
During the early stages of the fitting process, I used alignment of
the bulges as an initial guide towards a best fit; this seemed
reasonable since the bulges track the dynamical centres of the
galaxies. However, I soon revised the technique; the bulges do track the dynamical centres but their exact behaviour is uncertain under different kinds of galaxy models. Thus, I
decided to simply concentrate on using the tails to guide the matching
process.
Although I identified several lines of enquiry whose results would
prove useful for future model fitting, both circumstances and time
hampered any efforts to pursue them. In particular, I had hoped to
investigate how the inclination of the disks during encounter affected
results in terms of model dependency and also how changes in the
respective orbital inclinations of the individual galaxies would
change the results of the encounter.
Essentially, the primary purpose of my model matching efforts was not
so much to identify a best fit - unlikely given the size of the
parameter space - but to identify steps in the process which would
lend themselves to automation. Discussion with John would seem to
indicate that this primary goal was achieved.