SDSS-RF
B.2. COLOR-COLOR RELATIONS OF THE NEW MODELS 161
Figure B.9: Color vs. color plots for SDSS data (blue) and their individual best fittingSEDfit-SEDs (red) for 0.38≤z ≤ 0.42. Objects with uband errors lower than the median are indicated in light blue, whereas those with greater errors are dark blue. The mean error bars of all objects are shown in the lower right corners in dark blue, and the mean errors of objects withuband errors smaller than the median in light blue. The grid (upper panel) and the boundary (lower panel) within which the models are selected are shown in black. The dark green dots are the preselected models, whereas the light green points represent the models that are left over after the removal of redundant SEDs.
Appendix
C
SED Fitting Results for the BOSS CMASS Sample
In Chapter 6 we created models that should match the BOSS CMASS galaxies in terms of their colors in redshift bins with widths 0.04 centered onz = 0.5,0.6,0.7,0.8. We do so by fitting sets of model SEDs with different properties like metallicity, SFHs, and ages to the data (Sec. 6.2.1). In superposition to the SEDs, a burst model and additional dust extinctions for the burst and the main component are further degrees of freedom. In Figure C.1 we show the distributions of the χ2 values returned by the SED fitting code, when we fit the BC03, M05, and M11 models in equal setups to the data. The setup of the SED fitting and the code are detailed in Sec. 6.2.1 and are equal for all cases.
In Section 6.2.1 we modify the BC03 models, changing their continuum slope by multiplying witha·λ−β forλ≥λi with variousβ andλi values. This is done to introduce further degrees of freedom and to create models that match the colors of the data to a better extent. We fit the λ−β modifications of BC03 SSPs and CSPs with decreasing SFRs to the CMASS galaxy colors with the same setup as mentioned above and as detailed in Section 6.2.1. The resulting χ2 values of the fit are shown together with the previous results of the BC03, M05, and M11 models in Fig. C.1, where they are highlighted by orange histograms.
While the results of the original BC03 models, M05, and M11 produce similar results in terms of χ2, they are significantly improved by using the BC03 λ−β models. This is also shown in terms of colors in Figures 6.9 to 6.12 of Sec. 6.2.1.
10
-310
-210
-110
00.48 ≤z ≤0.52
2035 objects
BC03 BC03
λ−βM05 M11
10
-310
-210
-110
00.58 ≤z ≤0.62
2038 objects
0 1 2 3 4 5
10
-310
-210
-110
00.68 ≤z ≤0.72
2025 objects
0 1 2 3 4 5 10
-310
-210
-110
00.78 ≤z ≤0.82
1293 objects
χ2
no rm ali ze d f re qu enc y
Figure C.1: χ2 distributions of SED fitting runs (atzspec) detailed with BC03 (gray), BC03 λ−β (orange), M05 (blue), and M11 (green) models for the four considered redshift regions. The frequency is normalized and plotted logarithmically. We can see that although the distributions for BC03, M05, and M11 are similar, theχ2 values are improved by including the BC03λ−β models.
Appendix
D
SED Fitting Results for the Quintuply Lensed High-z System in RXC J2248
In Chapter 8 we estimated the physical properties of the high-z quintuply lensed system in RXC J2248 by SED fitting with SEDfit. We present here the results of the SED fitting in greater detail. We concentrate again only on ID2&3 which have the cleanest photometry.
Figures D.1 to D.4 display the results for exponentially increasing (τ < 0) and decreasing (τ > 0) SFRs. Figures D.5 D.6 show the fitting results with SSP models only, whereas Figures D.7 and D.8 display the results when we perform the SED fit to all models (SSPs and CSPs) combined. In the lower panels of Figures D.1 to D.8 we plot the best fitting SED and the photometry. The panels in the middle and upper rows show the 2-dimensional likelihood distributions of fitting parameters (Z, τ, model age,AV), whereas the probability distributions of the mass-to-light ratios (M/L) in theV band are shown in the upper panels.
The blue contours in the likelihood distributions outline the 1σ (solid), 2σ (dashed), and 3σ (dotted) confidence levels. The filter bands in which the S/N ratio does not exceed 1 (the dropout filters f225w to f775w) are considered upper limits in the SED fit (lower panels) and are highlighted by empty circles.
Figure D.9 shows furthermore the probability distributions of the model ages, marginalized over the other fitting parameters. The upper limits of the 2σ interval in age is smallest for the fit with SSP models (third panel of Fig. D.9). This is because all stars are assumed to form at the time of formation and therefore the mean stellar age of a SSP is higher than that of a CSP with the same formation redshift observed at the same cosmic time.
Figure D.1: SED fitting results for ID2 and models with negativeτ SFRs. The lower panel shows the input photometry and errors with black points. Empty circles denote the fluxes in the dropout filters where the fluxes are considered upper limits. The best fitting model SED is shown in blue and the convolved fluxes in the detection bands are displayed by filled circles. The density plots in the upper panels show the likelihood distributions of the SED fit in two-dimensional parameter spaces.
Blue lines denote the 1σ(solid), 2σ(dashed), and 3σ(dotted) confidence levels. Finally, the likelihood distribution of the mass-to-light ratio in the V band is plotted in the upper right panel. The dotted line in this panel denotes theM/Lratio of the best fitting model.
167
Figure D.2: SED fitting results for ID3 and models with negativeτSFRs. For a detailed description of the plot see Fig. D.1.
Figure D.3: SED fitting results for ID2 and models with positiveτSFRs. For a detailed description of the plot see Fig. D.1.
169
Figure D.4: SED fitting results for ID3 and models with positiveτSFRs. For a detailed description of the plot see Fig. D.1.
Figure D.5: SED fitting results for ID2 and SSP models. For a detailed description of the plot see Fig. D.1.
171
Figure D.6: SED fitting results for ID3 and SSP models. For a detailed description of the plot see Fig. D.1.
Figure D.7: SED fitting results for ID2 and all models (CSPs and SSPs). For a detailed description of the plot see Fig. D.1.
173
Figure D.8: SED fitting results for ID3 and all models (CSPs and SSPs). For a detailed description of the plot see Fig. D.1.
Figure D.9: PDF of the model age for ID2&3, marginalized over the other fitting parameters for a fit withτ <0,τ >0 CSPs, SSPs models, and for all models combined (from upper to lower panel).
The dashed lines are the ages of the best fitting models. Between the dotted lines the total probability is 95.45 % (corresponding to a 2σconfidence interval).
Appendix
E
SDSS CasJobs Queries
This chapter contains the queries used for the acquisition of the SDSS data from the SDSS database in order to give the reader more information about the catalogs and to provide the means to duplicate any SDSS catalogs used in this work. SDSS provides the Java program CasJobs on their database website1 which enables the user to write database queries in the Structured Query Language (SQL). If desired, queries can also be executed on the CasJobs website. These queries will be demonstrated in the following, where we will also give an explanation of the different data values acquired as well as the tables in the SDSS database from which they originate.