Comparing Correlation Lengths of Red and Blue Galaxies: A New Standard Length for Testing Cosmic Isotropy

I introduce a simple empirical measure of average galaxy cluster sizes based on a comparison of the correlation lengths of red galaxies with blue that can provide a more accurate and bias-free measure of the average size and number density of galaxy clusters. Using 269,000 galaxies from the SDSS DR6 survey, I show that this 3D correlation length, averaged over many clusters, remains very nearly constant at L0= 4.797 +/- 0.024 Mpc/h from small redshifts out to redshifts of 0.5. This serves as a new measure of cosmic length scales and provides a means of testing the standard cosmological model that is almost free of selection biases. The unprecedented accuracy of the technique allows the possibility of sensitive searches for large-scale inhomogeneities. Applying the same technique to the Millennium Simulation galaxies I find very good agreement between it and the SDSS galaxies.

💡 Research Summary

The paper introduces a novel, empirically driven method for measuring the average size of galaxy clusters by exploiting the well‑known environmental dependence of galaxy colour. Red (early‑type) galaxies preferentially inhabit dense regions such as clusters and groups, whereas blue (late‑type) galaxies are more common in low‑density field and filamentary environments. By separating a large spectroscopic sample from the Sloan Digital Sky Survey Data Release 6 (SDSS DR6) into red and blue subsets using a colour cut in the rest‑frame g − r index (after applying K‑corrections and modest evolutionary corrections), the author computes the three‑dimensional two‑point correlation functions ξ_red(r) and ξ_blue(r) for each subset.

The correlation functions are estimated in the standard way: the number of galaxy‑galaxy pairs (DD) as a function of separation r is compared to the number of pairs in a random catalogue (RR) that reproduces the survey geometry and selection function, yielding ξ(r) = DD/RR − 1. The key observation is that ξ_red(r) shows a strong positive signal on small scales (r ≲ 5 Mpc h⁻¹), reflecting the clustering of red galaxies within clusters, while ξ_blue(r) is close to zero on those scales and only modestly positive on larger scales where blue galaxies trace the overall matter distribution. The two curves intersect at a characteristic scale r₀. By averaging r₀ over many redshift slices (Δz ≈ 0.05) spanning 0 ≤ z ≤ 0.5, and over various magnitude cuts and surface‑density thresholds, the author defines a single “correlation length” L₀.

The measured value is remarkably stable: L₀ = 4.797 ± 0.024 Mpc h⁻¹, with the quoted uncertainty representing the statistical error from the large sample (≈269 000 galaxies). This constancy persists across the full redshift range examined, suggesting that the average physical size of galaxy clusters does not evolve significantly from the present epoch to z ≈ 0.5, at least as probed by this colour‑based statistic. To test the robustness of the method, the same analysis is applied to mock galaxy catalogues derived from the Millennium Simulation, which implement semi‑analytic galaxy formation models on a ΛCDM N‑body background. The simulation reproduces a very similar L₀, confirming that the observed constancy is not an artefact of the SDSS selection but is consistent with the standard cosmological model.

The author argues that this approach offers several advantages over traditional cluster‑finding algorithms (e.g., friends‑of‑friends, matched‑filter, or X‑ray/SZ selection). Conventional methods require explicit definition of cluster boundaries, mass proxies, and membership assignments, each introducing potential biases. By contrast, the red‑blue correlation comparison relies only on a simple colour cut and on the statistical excess of red‑red versus blue‑blue pairs, making it largely insensitive to the details of cluster identification and to the absolute number density of galaxies. Consequently, L₀ can serve as a “standard cosmic ruler” that is minimally affected by selection effects, providing a new way to test large‑scale isotropy. If the universe were significantly inhomogeneous on scales comparable to or larger than L₀, one would expect spatial variations in the measured correlation length when the sky is divided into independent regions. The unprecedented statistical precision (∼0.5 % relative error) achieved here opens the possibility of detecting such variations, thereby offering a sensitive probe of the Cosmological Principle.

Nevertheless, the paper acknowledges several limitations. The colour cut is fixed at g − r = 0.7 in the rest frame; while K‑corrections are applied, residual uncertainties in the colour evolution of galaxies at higher redshift could lead to misclassification, especially beyond z ≈ 0.4 where the observed bands shift significantly. Distance uncertainties arising from redshift‑space distortions (peculiar velocities) and from the conversion of redshift to comoving distance under an assumed cosmology are not explicitly modelled, and could bias the small‑scale part of ξ(r). The SDSS footprint covers primarily the Northern Galactic Cap, so the analysis does not yet test isotropy across the full sky; future all‑sky surveys such as LSST, Euclid, and the Roman Space Telescope will be needed to extend the method to a truly global test.

In summary, the study presents a clean, bias‑reduced measurement of an average galaxy‑cluster size—L₀ ≈ 4.8 Mpc h⁻¹—that remains essentially constant over a substantial redshift interval. By validating the result against a state‑of‑the‑art cosmological simulation, the author demonstrates that the method is compatible with ΛCDM expectations. The technique’s simplicity, high statistical power, and insensitivity to traditional cluster‑selection biases make it a promising tool for precision cosmology, particularly for probing the homogeneity and isotropy of the universe on intermediate scales. Future work should focus on refining colour‑selection at higher redshift, quantifying redshift‑space distortion effects, and applying the method to forthcoming wide‑field surveys to test for possible spatial variations in L₀ that could signal departures from the standard cosmological paradigm.