2014年7月28 (月) 15:00
Alan Lefor (Astronomical Institute, Tohoku University)
COMPARATIVE STUDIES OF STRONG GRAVITATIONAL LENS MODELS
Data from strong gravitational lensing is critically im-
portant in the era of precision cosmology. Analysis of
strong gravitational lensing depends on software analy-
sis of observational data.
Following a systematic review of 14 lens model soft- ware packages (Lefor et al. 2013), we studied the behav- ior of strong gravitational lens modeling software with changes in redshift (Lefor and Futamase 2013). Four dif- ferent strong gravitational lens codes are directly com- pared (Lenstool, glafic, GRALE and PixeLens) in the analysis of a mock model and SDSSJ1004+4112. The percent change in time delays calculated at each redshift tested are compared with percent change in Dd Ds /Dds . A mock model with a singular isothermal ellipsoid and four images is tested with each code. Five models are used with a constant zlens and a varying zsource , and five models with a constant zsource and a varying zlens . In general, the changes in time delay are of a similar mag- nitude and direction, although some calculated time de- lays did not follow changes in Dd Ds /Dds . This variation is explained by changes in image position calculated by glafic and GRALE, which varied according to Dds /Ds . Small changes in redshift affect the calculated time de- lay and mass, and that the effect on the calculations is dependent on the particular software used.
The next study directly compares the analysis of four known lens systems using four lens model software codes to understand the differences and limitations of the mod- els (Lefor and Futamase 2014). HydraLens 1 was used to generate multiple models for four strong lens systems in- cluding COSMOS J095930+023427, SDSS J1320+1644, SDSSJ1430+4105 and J1000+0021 (Lefor 2014). All four systems were modeled with PixeLens, Lenstool, glafic, and Lensmodel. The calculation of the Einstein radius and enclosed mass for each lens model was comparable. The results were more dissimilar if the masses of more than one lens potential were free-parameters. The image tracing algorithms of the software are different, result- ing in different output image positions and differences in time delay and magnification calculations, as well as el- lipticity and position angle. Differences in optimization resulted in different results for ellipticity and position an- gle. In a comparison of different software versions using identical model input files, results differed significantly when using two versions of the same software.
Taken together, these studies support the need for fu- ture lensing studies to include multiple lens models, use of open software, and availability of the lens model files used in studies.
【The Colloquium Committee in 2014】
Shijun Yoshida (yoshida<at>astr.tohoku.ac.jp)
Kimihiro Yamazaki (k.yamazaki<at>astr.tohoku.ac.jp)
Yusuke Okayasu (y.okayasu<at>astr.tohoku.ac.jp)