The supersymmetric (SUSY) extension of the standard model (SM) is one of the most attractive candidates for the physics beyond the SM. As well as solving the hierarchy problem to a great amount, the SUSY SM often predicts an apparently correct GUT unification. In addition, many SUSY models provide a dark matter (DM) candidate.Once the presence of SUSY is inferred from experiments, the next problem is to search for the underlying mechanism of SUSY breaking. The scale of SUSY breaking is one of the most important clue for this search.

An attractive possibility of SUSY breaking mechanism is the gauge mediation model because in addition to the conceptual simplicity of using the well-established field theory as a building block of model construction, the gauge mediation model has an advantage in solving the flavor-changing neutral current problem. In particular, the low-scale gauge mediation model with a light gravitino of mass m(3/2)≦O(10) eV is very attractive since there is no cosmological problem for such a model. The light gravitino scenario is also interesting in the aspect of experimental testability since they typically predict lighter SUSY particles, which may be in the energy region able to be tested at the LHC.

In the thesis, we first review various types of low-scale gauge mediation models and their theoretical and phenomenological aspects. The theoretical constraints together with the assumption of a light gravitino set strong restrictions on the models. For example, the model with a perturbatively stable SUSY breaking vacuum is already excluded by the Tevatron and LHC experiments. Other interesting possibilities are the minimal type model with a metastable vacuum and strongly interacting gauge mediation models. Strongly interacting models predict interesting possibility of lighter gauginos,but their prediction is rather ambiguous due to the strong interaction. In the thesis, we concentrate on the minimal model and study their collider phenomenology.

We review the experimental constraints and update the constraints coming from the recent results of the LHC experiments. We also show the discovery potential of the LHC in the near future. The result shows that the LHC is capable of testing a wide class of interesting models. In particular, if the number of messenger is less than 6, the minimal type gauge mediation model can be tested completely within a few 10 fb(-1) at the 14 TeV run.

In the main part of the thesis, we study the test of low-scale gauge mediation models after discovery. We particularly study possible test at earlier stages. We consider a method using significances in multiple modes. This method is applicable from the early stage and is useful because it automatically incorporates the possible contamination from the SM background and can be used for a quantitative analysis.

As a concrete example, we study the discrimination of the minimal gauge mediation model and the minimal gravity mediation model. These models have different scales of SUSY breaking and predicts a quite different physics in high energy and cosmological history. Collider phenomenologically, however, their signatures are similar. However,because of the fact that usually almost all the produced SUSY particles decay into the next-to-lightest SUSY particle (NLSP) before decaying into the gravitino and that in gravity mediation models it is not always true, the leptons or photons from the NLSP decay may be a discriminant of the two models. However, this discrimination is of a quantitative nature and we are interested in the case with a smaller number of signal events, the estimation of the SM backgrouds and detector efficiencies are of primary importance. We perform a Monte Carlo simulation and pay attention to these effects.We propose a discrimination based on using significance variables in multiple modes.Using the lepton, tau-jet and photon modes as well as the jets plus no lepton mode, we show that the discrimination is possible at a very early stage after discovery.

We also applied our method for the discrimination of the co-annihilation region and the focus point region in the minimal gravity mediation model. By using the significances in the two jets and four jets modes, we show that the discrimination also possible at an early time. The discrimination is very important because it gives an insight on the origin of the DM.

本論文は6章からなり第5章が主題である。第1章はイントロダクションであり、博士論文の背景、特に超対称性標準模型についての宇宙論および加速器物理からの現在の制限について解説してある。第2章は博士論文の主題である低スケールゲージ伝達模型を考える動機を議論し、さらにその模型の理論的詳細についても与えられている。第3章では模型から予言される超対称粒子の質量、それらの LHC 実験での生成断面積、および超対称粒子の崩壊幅が議論されている。第4章では第3章で与えられた予言をもとにLHC実験においてどのように低スケールゲージ伝達模型を検証するかが議論されている。解析の結果、超対称粒子生成に伴う終状態に含まれる高エネルギー光子対もしく高エネルギーレプトン対を探すことで低スケールゲージ伝達模型において興味のある質量領域の多くが LHC 実験 (14TeV, 10 fb-1) で発見可能であることが示された。

第5章では博士論文の主題である低スケールゲージ伝達模型の他の模型からの実験的識別について議論してある。本研究では模型同士を区別する計量として、各模型からシミュレーションを通して予言される数種類のシグナルに対する統計的有意性で張られる空間を考え、その空間内で適当に定義された模型間の距離を用いることを提案している。なお、統計的有意性で張られる空間の次元はシグナルの種類の数に一致する。本研究ではこの方法を用いて具体的にどの程度低スケールゲージ伝達模型が他の模型から区別可能かについてのデモンストレーションも試みられている。その結果ある模型の発見に必要なLHCのルミノシティーと同程度のルミノシティーでかなりの数の模型が区別可能であることが示され、提案された手法が模型の効率的かつ定量的な選別に役立つことが示された。

本研究に先行する模型間の実験的識別を議論する論文として Arkani-Hamed et.al. JHEP 0608 (2006) 070 が挙げられるが、その研究においては実際の実験データの解析に重要となる標準模型からの取り除けない背景事象の効果が取り入れられていない。一方、本研究では統計的優位性で張られる空間を考えているため、シグナルに対する標準模型からの背景事象を自動的に取り入れられており、この点において先行研究よりも優れている。

なお、本論文第5章は浅井氏(東大)・白井氏(バークレー)との共同研究であるが、論文提出者が主体となって解析を行ったもので論文提出者の寄与が十分であると判断する。

従って、博士(理学)の学位を授与できると認める。