To effectively separate extragalactic globular clusters from foreground stars and background galaxies requires both static deep photometry in a range of filters covering a wide wavelength range (since globular clusters lie between stars and galaxies in colour-colour space) and high spatial resolution since extragalactic globular clusters are more extended than foreground stars but more compact than background galaxies with the same apparent magnitude. The existing metrics that best capture the demands of extragalactic globular cluster science are the static science cosmology metric and galaxy counts metric.
To measure the properties of extragalactic globular clusters - their metallicities, ages and masses - also requires deep multiband photometry with a wide wavelength range.
The ideal survey strategy for extragalactic globular cluster studies would be the baseline 2.1 simulation (including the Virgo Cluster in the WFD footprint) but with an equal number of 50s u-band exposures. I have only answered questions that I have an opinion on. These comments were originally submitted in response for a request to feedback from the Galaxies collaboration but it was suggested that they be posted here as a more permanent record.
As a note methodology, to estimate the number of globular cluster Rubin would observe in each band, I used the WFD Extragalactic Coadded M5 median depths, a globular cluster luminosity function based on the Milky Way (a Gaussian with a mean of M_r = -7.7 which corresponds to a stellar mass of 2 x 105 solar masses and a sigma of 1.1 mag), colours based on an 12 Gyr [Fe/H] = -1 FSPS model and a globular cluster volume density based on the Driver et al. (2022) galaxy stellar mass function and the Eadie et al. (2022) globular cluster mass-galaxy stellar mass relation (this gives a volume density of about 6 globular clusters per cubic megaparsec).
Filter Distribution
1a Should the survey cadence skew towards bluer filter observations (compared to the current baseline)?
More g exposures provide little benefit in terms of the number of star clusters observed or how well we can constrain their properties; adding u exposures does. Deeper u-band photometry provides better mass, age and metallicity constraints, especially for younger and dustier clusters. However, the increase in depth from going from 1x30s to 1x50s with the same number of exposures (from u = 25.62 to u = 25.98) only provides modest improvements (5% to 20%) in the constraints. The number of globular clusters detected at 5 sigma in the u-band does dramatically increase by 60% though. Having u-band photometry is quite useful for separating globular clusters from foreground stars and background galaxies since globular clusters have redder (u - g) colours than galaxies with the same (r - z) colours but bluer (u - g) colours than stars with the same (r - z). Cantiello et al. (2020) saw a reduction in contaminates of 50% in their study of globular clusters in the Fornax cluster when using u-band photometry.
1c What is the exposure time for the u-band observations?
1x50s u band exposures provides a clear benefit in u-band depth, especially when the number of u band exposures is kept constant.
1e Should the survey use a different exposure than 2x15s (or possibly 1x30s) for non-u-band filters?
Longer visits lead to deeper coadded photometry for static sources such as star clusters but going to 40 s visits from 30 s does not provide a huge improvement. Shorter visits lead to shallower photometry.
Footprint
3a What should the exact Declination and dust extinction limits for the WFD region?
Overlap with Euclid should be maximised to take advantage of the superior spatial resolution of Euclid and Euclid’s near-infrared photometry. For extragalactic globular cluster science areas with low extinction should be favoured both since they allow deeper u and g band photometry but also since they have lower densities of foreground stars that have to be separated from extragalactic star clusters.
3b What should the definition of the Galactic bulge region be (e.g. should we add the Virgo cluster to WFD)
There are massive gains in the number of nearby (d < 20 Mpc) globular clusters by including the Virgo Cluster, the largest amount of stellar mass in the nearby universe.
3c How much time is spent observing the Galactic Plane?
The galactic plane area is less useful for extragalactic star cluster science due the high extinction and the high number of foreground stars. Extragalactic star cluster science would suffer if more time was added to the Galactic Plane.
3d How much time is spent observing the NES? and 3e How much time is spent covering the South Celestial Pole?
There is a trade off between more time for the NES and SCP allowing more star clusters to be studied in these areas and having fewer star clusters in the WFD. However, the NES and SCP will be observed at higher airmasses than the WFD which will lead to poorer effective seeing, limiting the ability to distinguish star clusters and background galaxies. The NES also lacks u-band photometry, limiting star-star cluster separation. Adding extra time to these areas would detract from extragalactic star cluster science.
Micro surveys
Local Volume Galaxies
These galaxies are close enough that star clusters are semi-resolved in Rubin imaging and thus can be separated from foreground stars by their extended morphologies, allowing for a particularly clean selection. Deeper photometry in g (and to a less extent r and i) will allow star clusters in these galaxies to be studied to much lower masses, allowing the processes that shape star cluster formation and evolution to be studied in depth. All 10 galaxies should get deeper photometry to allow the effects of galaxy morphology, star formation rate and mass on the population of low mass clusters to be understood. Even the smallest of the proposed changes (+50% g band exposures) provides a benefit.
Short exposures
Short exposures could prove useful in obtaining integrated photometry of massive star clusters in the Milky Way and its satellite galaxies. Having integrated photometry of star clusters that we know the ages, metallicities and masses of from resolved star studies observed with the same instrument and telescope provides a useful check on the measurements of the ages, metallicity and masses of extragalactic star clusters as well as measurements of masses, star formation histories and metallicities of galaxies observed with Rubin. Only a handful of exposures in each band would be required during the course of the survey.