Star formation activity in the Galactic center (GC) region is presently not impressive, at least outside a few special locations, first and foremost SgrB2. Still, the ISM in this region may be a nearby `laboratory for starburst galaxies' just as the rather quiescent SgrA^* has been called a laboratory for AGN (Mezger et al. 1997), sharing at least some characteristics with its far more active extragalactic cousins. Thus, understanding the unique properties of the ISM in the GC region is not only important in its own right but also has the potential to influence and challenge our conclusions about objects that are too far away to study at high spatial resolution and/or to observe in lines from rarer molecular species.
Since the main constituent of molecular clouds, H2, lacks a permanent dipole moment, molecular gas that is not hot enough to excite vibrational (infrared) H2 transitions -- only very exceptional clouds manage this -- has to be detected in `tracer molecules'. The most abundant of those, and thus the workhorse of molecular spectroscopists, is the main isotope of carbon monoxide, 12CO, with a relative abundance to H2 of ~10-4. Most maps of molecular gas distributions in the Milky Way and, even more pronounced, in external galaxies, are taken in the rotational J = 1 -> 0 transition of this molecule at 115GHz (2.6mm). CO also has the advantage of being easily excited even at low gas densities of only ~102cm-3. Thus, it traces almost all molecular gas.
Why isn't it, then, sufficient to survey the CMZ -- and external galaxies -- once in 12CO and be done with it? Inspection of the images obtained by surveys in a number of molecules and transitions readily provides the answer: The gas distribution looks very different, depending on what species and what excitation state is used. 12CO 1 -> 0 maps (see Bitran et al. and Jackson et al. 1996 for recent examples) tell us the morphology of molecular gas of all kinds -- and little else. It is striking how little contrast is seen in 12CO: These images let the observer question the concept of distinct molecular clouds in the GC. And this is, indeed, an important insight: 12CO picks up (among other things) a diffuse `intercloud medium', that, in the GC, is molecular rather than atomic. To get to the physical meat of the matter, the density and temperature distribution of the gas, multilevel studies, isotope studies and surveys in rarer, weaker molecular species are required. Since a field of at the very least 5o * 2o has to be covered, this makes for fairly tedious work, and it sometimes seems that more data have been accumulated for the nuclear regions of some external galaxies than for our GC.
As an example for the usefulness of such multiline surveys, let us consider the case of the `conversion factor': An analysis of line ratios clearly demonstrates the traps one can fall into if one considers only one line. A `standard' factor to convert integrated 12CO 1 -> 0 intensity to H2 column density has been established (by Strong et al. 1988 and in intense discussions in the subsequent years, for clouds in the disk of the Galaxy). This infamous quantity has been very widely used to determine gas masses in external galaxies, often in the central regions -- in many cases, out of desperation since measurements of lines other than 12CO 1 -> 0 are exceedingly difficult. In an analysis of line ratios, we (Dahmen et al. 1998) could show that not only has the `average' conversion factor to be lowered for the GC region by a factor of almost 10 -- the situation is even worse: There is no simple causal connection between 12CO 1 -> 0 line intensity and H2 column density. It is the otherwise convenient fact that the 12CO 1 -> 0 transition is so easily excited even at low gas densities that wreaks havoc here: A very significant amount of the total brightness of the emission arises from a thin, warm, unbound gas component that contributes little to the mass and does not exist in Galactic disk clouds. The presence of this component is tied to the special conditions in the GC region: large tidal forces and tri-axial (i.e. bar) potentials lead to shocks and cloud-cloud collisions that disrupt the material and leave part of it in a `diffuse', unbound state. We have every reason to believe that these processes also work in the centers of external galaxies -- only very desperate people should turn to `standard conversion' to determine gas masses in these regions anymore. It was only the combination of large scale surveys of the GC region, where the clouds can be resolved on pc scales, that made clear not only that `standard' CO-H2 conversion breaks down under Galactic center conditions, but also why this is the case.
Of course, molecular surveys are also used to trace the dynamics of the gas in the potential of the GC region, which is defined by the stellar component. Over the last years, a bar potential has become the by far most popular model to explain the unique kinematics (e.g. the large non-circular motions) observed. Again, surveys in different molecular species and transitions help to clarify matters, by pointing out regions of high and low density, helping to locate shocks and regions of peculiar chemistry.
More molecules? Molecules other than CO (CS, HCN, and -- on a slightly smaller scale -- SiO (Martin-Pintado et al. 1997) and HNCO (Dahmen et al. 1997) have already been surveyed) need higher gas densities to become excited and thus trace different components of the ISM. Some species, like SiO and HNCO, that are very rare or are seen only under very special conditions in confined region in the Galactic disk are widespread and abundant in the CMZ. It is, of course, tempting to imagine having at least medium resolution data of them all: the CO isotopomers for general gas properties, CS and HCN as `all purpose' tracers of gas at higher densities (>= 104cm-3), HCN's isomer HNC as a tracer of cool and quiescent gas, CH3OH as an indicator of hot, dense material, SiO, SO and SO2 for shocks, the radical CN to locate region of enhanced UV field, i.e. photodissociation regions, the molecular ion HCO+ to point out regions with high cosmic ray flux ... the list could go on for quite a while. And of course one can only really analyze the data if one has more than one transition for each molecule.
In the absence of a dedicated survey telescope in the ~10m class with an array receiver, this will clearly remain a dream for many years to come. However, one should keep in mind that for some external galaxies, e.g. the prototypical starburst NGC253, we are approaching a situation where most of the molecules above mentioned have been observed, many interferometrically at high resolution. To have a `baseline' for understanding these observations, further efforts in mapping the CMZ are certainly needed, even if the projects are time-consuming and, at least at first look, `routine'. As time passes, more data will become available -- till then, it will help the efforts of everyone working in this field if survey data are published timely and readily made available in digital form to the community (at least on request). A single one-line survey will provide only very limited new information -- pointing out for yet another time that the gas distribution is asymmetric, dynamically consistent with a bar and that the gas is, on average, rather dense and warm, is not very original anymore. The true value of new surveys will become apparent in an analysis in the context of what already exists.