Dependency of photo-z accuracy on filter definition of photometric surveys [IMA]

http://arxiv.org/abs/1706.09586


Accurate measurement of photometric redshift (photo-z) is crucial for photometric galaxy surveys, hence it is very important to study how the photo-z accuracy depends on the design of filter systems, and find the best filter parameters which can optimize photo-z fitting. Here we investigate this problem for the Chinese Space Station Optical Survey (CSS-OS) as an example, which covers wide wavelength range from near-ultraviolet to near-infrared bands. We use the COSMOS galaxy catalog to create mock flux data, whose galaxy redshift and magnitude distributions are similar to that of the CSS-OS. We compared different public photo-z spectral energy distribution (SED) template-fitting codes, and choose to use a modified LePhare code in the filter calibration process. The code exploits information in flux upper-limits, which has been proven to be effective in the photo-z measurement. We then investigate the effect of each filter passband and filter parameters on photo-z fitting. We find that the g, r, and i bands can significantly affect the photo-z accuracy and fraction of catastrophic redshift in the CSS-OS or any survey with similar magnitude limits, and the current CSS-OS filter design can optimize the photo-z fitting process.

Read this paper on arXiv…

Y. Cao, Y. Gong, X. Meng, et. al.
Fri, 30 Jun 17
32/65

Comments: 10 pages, 9 figures, 2 tables