Normalized to: Jayson, J.
[1]
oai:arXiv.org:1708.03068 [pdf] - 1853038
Appraisal of the BAO equations
Submitted: 2017-08-10, last modified: 2019-03-19
The commonly used isotropic baryon acoustic oscillation (BAO) equation is an
approximation derived from empirical and geometric arguments, and the
equivalent anisotropic BAO equations were written down by analogy. Using
fit-lines to CMB compatible solutions, {\Omega}_{m} and H_{0} values have been
derived for BAO studies without recourse to those equations, and have been
applied for their appraisal. The isotropic expression becomes problematic at
precision levels of ~1 percent or better, and at high redshift values of z = /
>1. Most revealing, the anisotropic equations,
D_{M}(z)/D_{M,fid}(z)={\alpha}_{\perp}r_{d}/r_{d,fid}, and
D_{H}(z)/D_{H,fid}(z)={\alpha}_{\parallel}r_{d}/r_{d,fid}, are invalid when
{\alpha} ~ 1, since under that condition,
D_{M}(z)/D_{M,fid}(z)=D_{H}(z)/D_{H,fid}(z)=r_{d}/r_{d,fid} ~ 1, and neither
equation can be satisfied with anisotropic data. (The ratios are respectively,
the angular distance, the inverse Hubble parameter, and the comoving acoustic
horizon, each divided by its fiducial value). {\alpha} can be driven towards
unity for any data set, e.g., by applying the derived {\Omega}_{m}, H_{0} pair
as the core of a second iteration fiducial parameter-set. Thus, the anisotropic
equations are untenable. Dissociated from the BAO equations, we have extracted
weighted mean values of {\Omega}_{mw}=0.299+/-0.011 and H_{0w}=68.6+/-0.7 km
s-1 Mpc-1 for eight uncorrelated BAO data sets.
[2]
oai:arXiv.org:1605.03037 [pdf] - 1403862
Application of USNO-B1.0 towards selecting objects with displaced blue
and red components
Submitted: 2016-05-10
We have conducted a feasibility study to determine the effectiveness of using
USNO-B1.0 data to preferentially detect objects with displaced red and blue
components. A procedure was developed to search catalogue entries for such
objects, which include M dwarfs paired with white dwarfs or with earlier
main-sequence stars, and galaxies with asymmetric colour distributions.
Residual differences between red and blue and infrared and blue scanned
emulsion images define vectors, which, when appropriately aligned and of
sufficient length, signal potential candidates. Test sample sets were analysed
to evaluate the effective discrimination of the technique. Over 91,000
USNO-B1.0 catalogue entries at points throughout the celestial sphere were then
filtered for acceptable combinations of entry observations and magnitudes and
the resulting total of about 17,000 entries was winnowed down to a little more
than 200 objects of interest. These were screened by visual examination of
photo images to a final total of 146 candidates. About one quarter of these
candidates coincide with SDSS data. Those constituents fall into two groups,
single and paired objects. SDSS identified several galaxies in the first group.
Regarding the second group, at least half of its members were tentatively
identified as main-sequence pairs, the greater portion being of widely
separated spectral types. Two white dwarf--main-sequence pairs were also
identified. Most importantly, the vectors formed from USNO-B1.0 residuals were
in alignment with corresponding SDSS pair position angles, thereby supporting
this work`s central thesis.