A SEARCH FOR WEAK DISTORTION OF DISTANT GALAXY IMAGES BY LARGE-SCALE STRUCTURE

We have completed a feasibility study for the measurement of weak distortion of distant galaxy images by intervening large-scale structure by using the 5-m Hale reflector to acquire a very deep, r approximately 26, exposure of a single field. The error budget of our observations is dominated by the effects of atmospheric seeing (which strongly degrades this signal because the faintest images are under-resolved) and telescope effects. After performing a correction for telescope aberrations and possible guiding errors, the observed mean 'polarization' of the images of 4363 galaxies with magnitudes 23 less-than-or-equal-to r less-than-or-equal-to 26 within a circle of radius 4.8 arcmin was found to be pBAR = 0.01 +/- 0.01. The associated two-point polarization correlation function has a constant value of C(pp) = (1.4 +/- 3.0) x 10(-5) over the angular range 1 to 6 arcmin. It is predicted that the cosmological polarization should be in the range p = 0.03 +/- 0.01 for a standard CDM universe normalized by a bias parameter, b, of unity (p scales inversely with b and approximately linearly with OMEGA0). For the atmospheric seeing and sky noise conditions associated with our data, Monte Carlo simulations suggest that the efficiency of measuring the mean cosmological polarization is on the order of 40 +/ 10 per cent. Thus our preliminary analysis suggests an upper limit on the cosmological mean polarization in the field of pBAR(max) approximately 0.04. Deep wide-field imaging in 0.5 arcsec seeing to study this polarization signal should provide limits that constrain current cosmological models.