Table 1: Simulation Parameters
D. R. Ingram and N. Katz
University of Washington, ASTRONOMY, Box 351580, Seattle, WA 98195-1580
D. H. Weinberg
Ohio State University, Department of Astronomy, Columbus, OH 43210
L. Hernquist
University of California, Lick Observatory, Lick Observatory, Santa Cruz, CA 95064
comoving box at high
redshifts. By converting these boxes into two-dimensional images,
convolving with a Gaussian beam and adding noise appropriate to
sensitivity estimates, we have simulated a series of observations
by the proposed Square Kilometer Array (SKA). The capability of
the SKA to easily detect 
/beam concentrations of
neutral Hydrogen (``galaxies") should impose significant constraints
upon the large scale structure proposed in a variety of cosmological
models.
One of the fundamental problems posed by observational cosmology over the last two decades is the origin of the large scale structure of the Universe, determined through extensive surveys in both the plane of the sky and in redshift space (e.g. Geller and Huchra 1989). Theoretical cosmology has presented us with a variety of proposed solutions to this problem in the form of competing cosmological models such as cold dark matter (CDM, see Peebles 1993) and models that incorporate a cosmological constant (Peebles 1993, Ratra and Peebles 1988). Recently, with the use of a new generation of computers, such as the Cray C-90, cosmological simulations have begun to attain levels of resolution and robustness sufficient to propose significant observational tests of cosmological model predictions.
The proposed Square Kilometer Array (SKA, see Braun 1995a) is a radio telescope that will be capable of groundbreaking work in observational cosmology. Using simulations based upon TreeSPH (Katz et al 1995, KWH hereafter), an algorithm that combines a hierarchical tree method (Barnes and Hut 1986) and smoothed-particle hydrodynamics (SPH, see Hernquist and Katz 1989 and references therein), we have produced three-dimensional maps of neutral Hydrogen in order to predict the 21 centimeter emission that will be observed by the SKA in its current proposed configuration.
The initial conditions for this simulation are listed in Table 1. Star formation (of necessity, a phenomenological algorithm due to particle sizes) was not included in this simulation. Unfortunately, the simulation runs that included star formation were not completed in time for presentation at this conference, but it is not believed that star formation will have a significant impact upon the general nature of the galaxy distribution (see KWH for a complete discussion).
Table 1: Simulation Parameters
Figure 1:
Simulated observation for 100 hours of integration time
by the Square Kilometer Array of neutral
Hydrogen. The image spans 
1 Mpc on a side at a redshift of 2.
Contours range from 10
to 5 x 10
/beam. The object at
Column 575, Line 500 is a ``galaxy" with roughly 10
.
SKA observations were simulated in the following way. Starting with a simulation data cube of neutral Hydrogen, we then selected an axis along which to view the box. The box was then divided into a number of velocity channels. In order to get a reasonable estimate of the sensitivity of the SKA in a single observation, we chose to look at frequency channel widths of 100 kHz, which corresponds to a velocity width of 65 km/s at a redshift of 2. This two-dimensional image of neutral Hydrogen was then converted into solar masses of Hydrogen per velocity channel (Braun 1995b):
where 
is solar masses of neutral Hydrogen per pixel, 
is
column density of neutral Hydrogen, S is the pixel size in centimeters
and 
is the mass of a Hydrogen atom.
Figure 2:
Simulated observation for 10000 hours of integration time
by the Square Kilometer Array of neutral
Hydrogen. The image spans 
1 Mpc on a side at a redshift of 2.
Contours are the same as in Figure 1. Objects of order
in neutral Hydrogen are now clearly seen.
This image is then convolved with a 35 kpc FWHM Gaussian to represent
the image of the sky seen by the SKA beam. After normalization, we
are left with an image in units of 
per beam. Using sensitivity
estimates from Braun (1995b), we then create a Gaussian noise image
with magnitude proportional to (exposure time)
. This
noise image is convolved with a 35 kpc Gaussian (for z = 2), normalized
and added to the neutral Hydrogen image. The resultant image is
a simulation of data collected in a single observation of the SKA of
a given field for a given exposure time.
Figure 1 shows
a contour plot of the central region of
such an image, for a 100 hour observation of a
representative velocity channel of the simulation box.
Figure 2 shots a similar contour plot, but for 
hours of
observation time.
As a rule of thumb, we find that the predicted capabilities of the SKA
should detect a 
/beam
concentration of neutral Hydrogen
at the 5
level in 100 hours. This means that, for the first time,
``normal" galaxies will be easily detectable at high redshifts, and
the information contained in the general properties of these galaxies
(such as the mass distribution and correlation function)
should strongly constrain theories of structure formation
and cosmological models.
Filamentary structures that could give rise to Lyman-alpha forest
spectral features
unfortunately have brightnesses
far too low (of order 
/beam and lower) to be seen, but
with the constraints on properties of normal galaxies that
should come from the SKA, we should have a much clearer picture
of the nature of the formation of large scale structure in the Universe.
In the coming months, we intend to redo the simulations under a variety of different cosmological models in the hopes of providing realistic observational tests for new instruments such as the Square Kilometer Array. There is also an excellent discusssion on the implications of current and future simulations by Weinberg (1995).
This document was generated using the LaTeX2HTML translator Version 95.1 (Fri Jan 20 1995) Copyright © 1993, 1994, Nikos Drakos, Computer Based Learning Unit, University of Leeds.
The command line arguments were:
latex2html -split 0 coldgas.tex.
The translation was initiated by Joanne Hughes on Tue Dec 19 11:15:51 PST 1995