利用Subaru超廣角可見光相機觀測資料模擬星系目錄

Abstract

人造星系目錄是一種模擬真實星系的統計性質進而創造出來的星系 目錄。此類目錄擁有巨量的星系樣本數以及涵蓋寬廣的觀測面積和足 夠的紅移範圍，因而能夠用於研究測量星系性質時所隱含的系統性誤 差或是選擇性偏差。基於上述，我們根據不同的目的呈現了兩種不同 的人造星系目錄。第一種星系目錄的用途在於測試交叉比對法，第二 種星系目錄的用途則在於利用多頻段光度人造星系目錄測試星系團偵 測演算法。 對於第一種人造星系目錄，我們利用公開可使用的 MICE 人造星系 目錄建立一個適用於 HSC 多頻段光度系統的模擬星系目錄。此星系目 錄所擁有多頻段光度資料有助於我們測試任何需要光譜能量分佈資訊 的分析方法。在本論文中，我們利用此星系目錄驗證一種統計推導星 系恆星質量函數的方法，並證實此方法所得的星系恆星質量函數相符 於真實的函數。 對於第二種人造星系目錄，我們利用實際觀測所得的星系目錄和來 自多體模擬的暗物質暈目錄建構一種特別用於測試星系團偵測演算法 的模擬星系目錄。藉由執行演算法於這類星系目錄，我們能夠量化分 析演算法所得結果進而推導初始條件與最終結果的相關性。我們發現 演算法所偵測到的星系團幾乎為我們所模擬的星系團。然而，並非所 有模擬的星系團皆被偵測到。我們發現模擬星系團能否被偵測到與其 紅序帶星系的比例及其暗物質暈的質量有明顯的相關性。

The galaxy cluster sciences have been flourishing in modern times due to the successful wide field surveys, such as the Sloan Digital Sky Survey (SDSS), the Hyper Suprime-Cam Subaru Strategic Program (HSC) and the Dark Energy Survey (DES). To study physical properties of these massive gravitational systems statistically, mock catalogs are an useful tool for evaluating statistical uncertainties in measurements and quantifying the observational selection bias. We present two types of mock catalogs designed for different purposes in the HSC cluster survey. One is for quantitatively studying the completeness and purity of the cluster sample, and the other is for statistical analysis. 1. To validate cross correlation methods, we construct a general photometric mock galaxy catalog based on the MICE-Grand Challenge Galaxy Catalog generated by combining Halo Occupation Distribution (HOD) method with Sub-halo Abundance Matching (SHAM). For the original MICE galaxy catalog, it covers an area over 5000 deg2 out to redshift z < 1.4 and is complete down to an intrinsic magnitude Mr < −18.9. We extract a patch of a 200 deg2 area and derive physical properties of galaxies, such as stellar mass and photometric redshift, from the DES and the Vista Hemisphere Survey (VHS) photometry originally provided by the MICE catalog. We also include observational artifacts, like bright star masks and irregular survey boundary. 2. For second type of mocks, we present a set of mock galaxy catalogs designed for cluster detection in the HSC survey. The catalogs are generated using real galaxies identified in the S16A release of HSC survey and a halo catalog from an N-body simulation. Our mock clusters span the redshift range from 0.3 to 1.2 with the mass range 1013M⊙ < M200 < 2 × 1014M⊙. To mimic realistic observation conditions, we make use of the HSC galaxy catalog in the COSMOS field as the field population and combine it with our mock cluster galaxies, adopting masked regions where no galaxies are in them. The CAMIRA algorithm is run on 90 such sets of mock catalogs. By cross-matching the result with halo catalogs, we investigate the purity and completeness of CAMIRA detected mock clusters, and find that the purity of the CAMIRA mock clusters is greater than 0.95, which indicates that most of the clusters detected by CAMIRA are true clusters in mocks. For the completeness, we find that it is correlated with the halo mass and redshift. The completeness is high at the high mass end and starts to drop once halo masses are lower than 1.4 × 1014M⊙. Given a halo mass, the completeness is higher at lower redshift and becomes lower at higher redshift.