| Three themes are involved in this thesis:1. Redshift-Space Distortions on Baryon Acous-tic Oscillations.2. Spectroscopic Target Selection for Extragalactic Survey.3. Design and Implementation of Job Scheduling of Virtual Observatory Data Access System (VO-DAS). The first two are the astronomical research, while the last one is intended to create a data-intensive platform via the existing network, which belongs to the astronomical techniques and methods.The redshift distortion is an observational effect. Due to the peculiar velocity of galaxies, the distance estimation via the Hubble theorem is biased, and will directly affect the galaxy distribution. At large scales, the distortion presents as a linear squashed effect. While at small scales, it elongates the pattern in the opposite way, giving rise to the fingers-of-God, which is a nonlinear effect. However, how do these two kinds of distortion affect the statistical analysis of galaxies distribution respectively? This question, which will be investigated in Chapter 3, is the first one of our three focuses in this thesis. Firstly, the linear theory of the redshift space correlations, considering three cases (in the line of sight(LOS), in the perpendicular direction and the angular average), predicts that:1. the linear effect from redshift distortion could amplify and sharpen the Baryon Acoustic Os-cillations(BAO) feature near the LOS direction, due to the compression of redshift space.2. The sharpening is much weaker far from the LOS, in the transverse (rp) direction.3. The redshift space compression is the overall smooth shift towards negative values along the LOS, and towards positive values in the transverse.That means the linear effect is beneficial to the BAO measurement, especially in the LOS direc-tion. Secondly, In order to test our predictions, we employed two simulated samples in order to check the predicts from the theory. One sample is from Millennium Simulation(MS), which covers a smaller volume(box size is just 500Mpc/h), but with a much higher resolution(about 2Mpc/h), the other one are from 100 low-resolution(4Mpc/h) realizations of a 1Gpc/h box. By cutting the samples into many 2D slices, we estimate the 2D correlation function in real space, in redshift space with linear effect, and in redshift space with both linear and nonlinear effect. The simulations tell us that the predicted sharpening and amplifying effect indeed exist in the linear case, but meanwhile the nonlinear effect blurs all the feature in the correlation function, even far from the LOS. That means BAO detection could be improved if the nonlinear effects(i.e. the fingers of god) are reliably removed, but other (i.e. linear) distortions are kept in place. Indeed, the error bars for the LOS BAO bump are slightly reduced compared with the angle-averaged bump. At last, we analyze the 660 2.5deg thick slices from DR7 Main Galaxy Sample(MGS), and see a bump along the line of sight that is sharper than one might expect. Using Mexican-hat wavelets, we assess its significance within the sample at about 4.0σ.Because the above adopted small sample (SDSS MGS) is too small after all, the uncertainty of BAO detection is a bit large. As we know the error from the analysis of galaxy clustering is inverse proportional to the square root of sample volume. More accurate measurement is driving the much larger survey nowadays, like the LAMOST redshift survey in China. For the redshift survey, the spectroscopic target must be selected before the survey starts. Two methods to select the targets are discussed in this thesis, which involved 4000A and Layman break techniques. Through stellar population synthesis, I simulated 700 spectra for seven types of Hubble galaxies , and find that much older the galaxy is, more obvious the 4000A break is, while the Layman break is opposite, especially in the E/SO type, the Layman break almost disappear. On the contrary, much younger the galaxy is, much stronger the Layman break is, much weaker the 4000A break is. But the 4000A break is still visible in most of younger galaxies, even in the Im type case. Therefore the 4000A break technique is better method than Layman break for the early-type galaxy such as the Luminous Red Galaxy(LRG); while Layman break technique is better than 4000A break for the late-type galaxy like Emission Line Galaxy(ELG). But in some cases the Lyman break can not be used, the 4000Abreak is still available for the target selection. LRG and ELG are generally considered as much ideal tracers for the large scale structure of universe, we have already selected several trial LRG and ELG samples for LAMOST project, which is still in commissioning stage.With the technological advances in design and manufacture of telescopes and instruments on the multi-band, the capability of astronomical observation is enhanced greatly, which brings the global astronomy to enter into data abundant era. Enough astronomical data is very im-portant for the astronomical research, especially for the Large Scale Structure(LSS) under the requirement of the precise cosmology. However, the avalanche of observed data also brings the challenge to the ability of data storage, access, analysis, etc.The Virtual Observatory(VO) appeared in the time when the astronomical community were trying to integrate global astronomical resources so that isolated data can be assembled at the fingertips of astronomers. In order to archive this goal, the global experts working on astronomy and computer science are making great efforts on this international project, so does the team from China-VO. Based on trials in the several years, the China-VO team designed and implemented a set of their own programme, namely the Virtual Observatory Data Access System (VO-DAS). The third part of this thesis presents mainly the job scheduling style of this system, in the view of the working flow, design pattern, session philosophy, lifecycle, the destroy of resourse and exception process, etc. modules. At last, several applications of this system on the astronomy science are given.
|
PDF Full Text Request