魏俊杰

(中國(guó)科學(xué)院紫金山天文臺(tái) 南京 210008)

博士學(xué)位論文摘要選登

不同天文觀測(cè)對(duì)宇宙學(xué)模型的限制

魏俊杰?

(中國(guó)科學(xué)院紫金山天文臺(tái) 南京 210008)

1998年,科學(xué)家們通過對(duì)Ia型超新星的觀測(cè)研究,發(fā)現(xiàn)了宇宙在加速膨脹,揭示了暗能量的存在.該發(fā)現(xiàn)使得宇宙學(xué)成為當(dāng)今物理學(xué)的研究熱點(diǎn).宇宙學(xué)是一門高度依賴于觀測(cè)事實(shí)的學(xué)科,利用各類天文觀測(cè)數(shù)據(jù)來限制不同的宇宙學(xué)理論模型是現(xiàn)代宇宙學(xué)的一項(xiàng)重要研究工作.

本論文的主要研究?jī)?nèi)容是利用不同的天文觀測(cè)來限制宇宙學(xué).這些天文觀測(cè)包括Ia型超新星、超亮的Ic型超新星、伽瑪射線暴、強(qiáng)引力透鏡、星系團(tuán)的角直徑距離以及星系年齡.本文第1章簡(jiǎn)要回顧了宇宙學(xué)的研究背景,介紹了各種宇宙學(xué)模型,并詳細(xì)總結(jié)了不同天文觀測(cè)數(shù)據(jù)在宇宙學(xué)方面的研究進(jìn)展.

第2章全面介紹了我們?cè)诔滦怯钪鎸W(xué)方面的研究.Ia型超新星作為標(biāo)準(zhǔn)燭光的主要困難是,我們必須對(duì)表征超新星光度的3–4個(gè)“討厭”參數(shù)連同宇宙膨脹模型的參量同時(shí)做擬合.對(duì)于擬合方法,我們認(rèn)為應(yīng)該采用最大似然估計(jì)法來擬合參數(shù)里含有未知內(nèi)稟誤差的模型.然而,傳統(tǒng)的擬合方法是通過調(diào)節(jié)內(nèi)稟誤差的大小使得約化的χ2等于1,進(jìn)而求出內(nèi)稟誤差的值.這種方法并沒有考慮參數(shù)之間可能存在的方差.我們利用超新星信息巡天樣本里的252顆超新星比較標(biāo)準(zhǔn)的ΛCDM模型和Rh=ct模型,發(fā)現(xiàn)每個(gè)模型都能夠很好地?cái)M合超新星數(shù)據(jù).另外,現(xiàn)有觀測(cè)證據(jù)已表明超亮的Ic型超新星也可作為宇宙學(xué)探針,它們甚至可以比Ia型超新星測(cè)量紅移更遠(yuǎn)(z?2)的宇宙.然而,目前可用的Ic型超新星樣本還是很少.我們的模擬表明如果將來觀測(cè)樣本不斷增多的話,超亮的Ic型超新星有望成為高效的宇宙學(xué)探針,它們可對(duì)宇宙學(xué)參數(shù)做出精確限制,特別是暗能量的狀態(tài)方程參數(shù)wde.

第3章我們則專注于伽瑪暴宇宙學(xué)的研究.首先,我們利用伽瑪暴的光度關(guān)系來構(gòu)造高紅移處的哈勃圖.目前的Ia型超新星觀測(cè)只能測(cè)量低紅移的宇宙,我們需要用高紅移的伽瑪暴來延伸哈勃圖.其次,我們利用伽瑪暴來限制高紅移的恒星形成率.我們利用最新的Sw ift觀測(cè)證實(shí)了高紅移處的伽瑪暴爆發(fā)率仍然超出恒星形成率的預(yù)期,伽瑪暴爆發(fā)率與恒星形成率之間的不一致可以用宇宙金屬豐度演化來解釋.假定伽瑪暴爆發(fā)率與恒星形成率和金屬豐度演化有關(guān),我們發(fā)現(xiàn)伽瑪暴數(shù)據(jù)對(duì)高紅移恒星形成率的斜率的限制為?2.41.此外,只有暗物質(zhì)暈結(jié)構(gòu)中的重子氣體密度足夠高時(shí),重子氣體才容易冷卻并形成恒星.而且重子氣體密度應(yīng)該是正比于暗物質(zhì)密度,因此恒星只能在大于一定質(zhì)量的暗暈結(jié)構(gòu)中形成.我們可用最小的暗暈質(zhì)量Mmin來表征恒星開始形成.Mmin在恒星形成過程中起到了關(guān)鍵性作用.在68%的置信度區(qū)間內(nèi),伽瑪暴數(shù)據(jù)對(duì)Mmin的限制為Mmin＜1012.5M⊙.

第4章我們收集了69個(gè)強(qiáng)引力透鏡系統(tǒng)的觀測(cè)數(shù)據(jù).強(qiáng)引力透鏡在宇宙學(xué)研究中所發(fā)揮的作用日漸凸顯,我們仔細(xì)描述了如何用這類數(shù)據(jù)來限制宇宙學(xué)參數(shù).盡管強(qiáng)引力透鏡數(shù)據(jù)的觀測(cè)精度還沒有好到可以衡量哪個(gè)宇宙學(xué)模型更優(yōu),但是我們發(fā)現(xiàn)標(biāo)準(zhǔn)的ΛCDM模型和Rh=ct模型都可以很好地解釋目前這些觀測(cè)數(shù)據(jù).

在第5和第6章中,我們分別利用星系團(tuán)的角直徑距離數(shù)據(jù)和32個(gè)古老星系的年齡測(cè)量來檢驗(yàn)和比較標(biāo)準(zhǔn)的ΛCDM模型和Rh=ct模型,我們發(fā)現(xiàn)每個(gè)模型都能夠很好地解釋這兩組數(shù)據(jù).然而,這兩個(gè)模型各自所含的自由參量個(gè)數(shù)不同,因此我們需要借助模型選擇標(biāo)準(zhǔn)來判斷哪個(gè)宇宙學(xué)模型的擬合優(yōu)度更高.Akaike、Ku llback和Bayes信息選擇標(biāo)準(zhǔn)表明,相比ΛCDM模型,這些觀測(cè)數(shù)據(jù)更加傾向于支持Rh=ct模型,似然概率分別約為70%,75%和80%.

最后,第7章是作者在完成了以上研究工作之后對(duì)現(xiàn)階段宇宙學(xué)研究中所存在問題的一點(diǎn)思考,以及對(duì)未來宇宙學(xué)研究前景的一些展望.

with the observations of Type Ia supernovae(SNe Ia),scientists discovered that the Universe is experiencing an accelerated expansion,and then revealed the existence of dark energy in 1998.Since the am azing discovery,cosm ology has became a hot topic in the physical research field.Cosm ology is a sub ject that strongly depends on the astronom ical observations.Therefore,constraining different cosmologicalmodels with all kinds of observations is one of themost important research works in themodern cosmology.

The goal of this thesis is to investigate cosm ology using the latest observations.The observations include SNe Ia,Type Ic Super Lum inous supernovae(SLSN Ic),Gamma-ray bursts(GRBs),angular diameter distanceofgalaxy cluster,strong gravitational lensing,and agem easurem ents of old passive galaxies,etc.In Chapter 1,we briefly review the research background of cosm ology,and introduce som e cosm ologicalm odels.Then we summarize the progress on cosmology from all kinds of observations in more details.

In Chapter 2,we present the results of our studies on the supernova cosmology.The m ain difficulty with the use of SNe Ia as standard cand les is that onemust optim ize three or four nuisance parameters characterizing SN lum inositiessimultaneously with the parameters ofan expansionmodelof the Universe.Wehave confirmed that one should optim izeallof the parameters by carrying out themethod ofm aximum likelihood estim ation in any situation where the param eters include an unknown intrinsic dispersion.The commonly used m ethod, which estimates the dispersion by requiring the reducedχ2to equal unity,does not take into account all possible variances among the parameters.We carry out such a com parison of the standardΛCDM cosmology and the Rh=ct Universe using the SN Legacy Survey sam ple of 252 SN events,and show that each model fits its individually reduced data very well.Moreover,it isquiteevident that SLSNe Icmay beuseful cosmologicalprobes,perhaps even out to redshiftsmuch greater(z?2)than those accessible using SNe Ia.However, the currently available sam p le of SNe Ia is still quite small.Our simulations have shown that if SLSNe Ic can be commonly detected in the future,they have the potentialof greatly refining themeasurement of cosmological parameters,particularly the parameter wdeof the dark energy equation of state.

In Chapter 3,we focus on GRB cosmology.We firstly use GRBs as standard cand les in constructing the Hubble diagram at redshifts beyond the current reach of SNe Ia observations.Then wem easure high-z star form ation rate(SFR)using GRBs.We con firm thatthe latest Sw ift sample of GRBs reveals an increasing evolution in the GRB rate relative to SFR at high redshifts.The observed discrepancy between the GRB rate and the SFR may be elim inated by assum ing a cosm ic evolution in metallicity.Assum ing that the SFR and GRB rate are related via an evolvingmetallicity,we find that the GRB data constrain the slope of the high-z SFR to be?2.41.In addition,first stars can only form in structures that are suitably dense,which can be param eterized by them inimum dark matter halomass Mmin.Mminmust p lay an important role in star formation.We can constrain Mmin＜1012.5M⊙a(bǔ)t 68%confidence level from the GRB data.

In Chapter 4,we assemble a catalog of 69 strong gravitational lensing system s,and carefully introduce how to constrain cosmological parameters using these im portant data. We find that bothΛCDM and the Rh=ct Universe account for the lens observations quite well,though the precision of these m easurem ents does not appear to be good enough tofavor onem odel over the other.

In Chapters 5 and 6,we use measurements of the galaxy-cluster angular diameter distances and 32 agemeasurements of passively evolving galaxies to test and com pare the standard m odel(ΛCDM)and the Rh=ct Universe,respectively.We show that bothmodels appear to account for these two data very well.However,because of the different number of free parameters in thesemodels,we have to judge the goodness-of-fit of cosmologicalmodels with selection tools,such as the Akaike,Ku llback,and Bayes Inform ation Criteria,favoring Rh=ct overΛCDM with a likelihood of about 70%,75%,and 80%,respectively.

Finally,some open questions and an outlook in the cosmology field are summarized in Chapter 7.

Constrain ing Cosm o logical M odels with D ifferent Observations

WEIJun-jie

(Pu rp le M oun tain O bserva to ry,Chinese A cadem y of Sciences,Nan jing 210008)

10.15940/j.cnki.0001-5245.2016.04.012

?2015-07-01獲得博士學(xué)位,導(dǎo)師:紫金山天文臺(tái)吳雪峰研究員、美國(guó)亞利桑那大學(xué)Fu lvio Melia教授;jjwei@pm o.ac.cn