TG-AST150012

NameTG-AST150012
DescriptionWe request a startup allocation to support development on two related projects. The main idea is to use three dimensional simulations to constrain the histories of observed galaxies. The majority of work (1) during the initial periods of this startup will be for Graduate Student S. Alireza Mortazavi to shift an existing Condor-based pipeline at the Space Telescope Science Institute (STScI) to the Open Science Grid in order to test and plan for an ambitious expansion of his research program to constrain the dynamical histories of galaxies (Mortazavi et al. 2015). PI Snyder will also begin to develop pipelines to exploit large-scale cosmological hydrodynamical simulations (2), which predict the evolution of entire populations of galaxies in representative model universes, requiring data-intensive computing. 1. Modeling the Initial Conditions of Interacting Galaxy Pairs Using Identikit We use the Identikit software (Barnes & Hibbard 2009, Barnes 2011; http://www.ifa.hawaii.edu/~barnes/research/identikit/ ) to model the dynamics of interacting galaxy pairs. By measuring the initial conditions of galaxy mergers, we can constrain both cosmology and galaxy astrophysics. A galactic encounter has several free parameters and it is time consuming to find the best match between model and data. However, Identikit combines multiple techniques to quickly explore parameter space to find the simulation most similar to the observed shape and constituent velocities. We have developed an automated pipeline based on the latest version of Identikit to scan parameter space and find robust matches and associated uncertainties (Mortazavi et al. 2015), implemented in an STScI-based Condor environment. We will continue to test our method against simulations of galaxy mergers to determine the systematic errors in our measurements. In addition, we will apply it to real data: We have observed a sample of ~30 interacting galaxy pairs using different telescopes. We have reached the limits of the HTCondor cluster at STScI, and therefore we seek to test the options available through XSEDE. We need around 50,000 SUs to compute matches and uncertainties of the measurements for each merging pair (observed or simulated), and so we are requesting a startup allocation of 150,000 SUs to perform tests while planning for a larger research allocation. This research will have direct applications for interpreting data from the Sloan Digital Sky Survey-IV survey "Mapping Nearby Galaxies at APO" (MaNGA). 2. Mock data applications from large hydrodynamical simulations PI Snyder will develop methods for converting large cosmological simulations of galaxy formation (e.g., the Illustris Project www.illustris-project.org) into direct predictions for astronomical observatories. With XSEDE resources, I will seek to expand our Mock Galaxy Observatory efforts in new and ambitious directions, such as creating synthetic survey fields and advanced spectroscopic data products. For instance, we will explore the possibility of using large cosmological simulations as benchmarks for the Identikit modeling described in project 1. For testing, we are requesting 50,000 SUs on Gordon, and Data Oasis storage of 5000GB, enough to store two Illustris Simulation (or similar) outputs plus post-processed data products. In future allocation requests, we may seek to make these model archives available to the community through an XSEDE Gateway.
OrganizationSpace Telescope Science Institute
DepartmentUnknown
Sponsor Campus GridOSG-XSEDE
Principal Investigator
Gregory Snyder
Field Of ScienceMathematical Sciences
DisableFalse