서브메뉴
검색
Sensitivity Analysis of Solid Earth Processes Using Reduced-Order Models
Sensitivity Analysis of Solid Earth Processes Using Reduced-Order Models
상세정보
- 자료유형
- 학위논문 서양
- 최종처리일시
- 20260202105300
- ISBN
- 9798270247881
- DDC
- 550
- 서명/저자
- Sensitivity Analysis of Solid Earth Processes Using Reduced-Order Models
- 발행사항
- [Sl] : University of California, San Diego, 2025
- 발행사항
- Ann Arbor : ProQuest Dissertations & Theses, 2025
- 형태사항
- 260 p
- 주기사항
- Source: Dissertations Abstracts International, Volume: 87-06, Section: B.
- 주기사항
- Advisor: May, Dave A.
- 학위논문주기
- Thesis (Ph.D.)--University of California, San Diego, 2025.
- 초록/해제
- 요약Numerical models used to understand subduction zones and earthquake rupture are subject to uncertainty because the Earth's interior cannot be directly observed. Robust sensitivity analysis can quantify the influence of model parameter variability on model outcomes, but this is challenging due to the computational expense of the models combined with the high number of parameters of interest. To address this challenge, we present a framework for model order reduction in geophysical applications. We use the interpolated Proper Orthogonal Decomposition to build fully data-driven, non-intrusive reduced order models (ROMs). We build ROMs that accurately approximate output from subduction zone thermal models and earthquake dynamic rupture models and are orders of magnitude faster to evaluate. We present ROMs for 2D kinematic-dynamic models of subduction zone temperature, and use them to quantify the variability in slab interface temperatures that results from model input variability and different modeling assumptions. We find that model input variability results in significant variability in temperature and the inferred potential rupture extent for megathrust earthquakes at the Cascadia, Nankai and Hikurangi subduction zones. Moving to 3D, we build ROMs for 3D dynamic models of subduction and mantle flow, approximating viscosity, velocity and the second strain rate invariant in the Cascadia subduction zone. This study demonstrates that ROMs are effective tools that enable sensitivity analysis for high-dimensional 3D geodynamic models. Finally, we present a combined mesh morphing and model order reduction framework that allows us to quantify model sensitivity to geometric variability. We apply this method to meshes for 2D subduction zone thermal models, incorporating realistic slab interface curvature and depth uncertainties, and to 3D earthquake dynamic rupture models, varying fault dip angles across a 40° range. The mesh morphing method enables us to build ROMs that incorporate geometric variability, and we demonstrate this with ROMs built from dynamic rupture output that predict surface displacement and velocity time series for variable fault geometries. Together, these contributions provide a framework for robust sensitivity analysis of high-dimensional simulations for a range of geophysics problems.
- 일반주제명
- Geophysics
- 일반주제명
- Soil sciences
- 키워드
- Dynamic rupture
- 키워드
- Subduction zones
- 기타저자
- University of California, San Diego Scripps Institution of Oceanography
- 기본자료저록
- Dissertations Abstracts International. 87-06B.
- 전자적 위치 및 접속
- 로그인 후 원문을 볼 수 있습니다.
MARC
008260126s2025 us c eng d■001000017360078
■00520260202105300
■006m o d
■007cr#unu||||||||
■020 ▼a9798270247881
■035 ▼a(MiAaPQ)AAI32281218
■040 ▼aMiAaPQ▼cMiAaPQ
■0820 ▼a550
■1001 ▼aHobson, Gabrielle M.
■24510▼aSensitivity Analysis of Solid Earth Processes Using Reduced-Order Models
■260 ▼a[Sl]▼bUniversity of California, San Diego▼c2025
■260 1▼aAnn Arbor▼bProQuest Dissertations & Theses▼c2025
■300 ▼a260 p
■500 ▼aSource: Dissertations Abstracts International, Volume: 87-06, Section: B.
■500 ▼aAdvisor: May, Dave A.
■5021 ▼aThesis (Ph.D.)--University of California, San Diego, 2025.
■520 ▼aNumerical models used to understand subduction zones and earthquake rupture are subject to uncertainty because the Earth's interior cannot be directly observed. Robust sensitivity analysis can quantify the influence of model parameter variability on model outcomes, but this is challenging due to the computational expense of the models combined with the high number of parameters of interest. To address this challenge, we present a framework for model order reduction in geophysical applications. We use the interpolated Proper Orthogonal Decomposition to build fully data-driven, non-intrusive reduced order models (ROMs). We build ROMs that accurately approximate output from subduction zone thermal models and earthquake dynamic rupture models and are orders of magnitude faster to evaluate. We present ROMs for 2D kinematic-dynamic models of subduction zone temperature, and use them to quantify the variability in slab interface temperatures that results from model input variability and different modeling assumptions. We find that model input variability results in significant variability in temperature and the inferred potential rupture extent for megathrust earthquakes at the Cascadia, Nankai and Hikurangi subduction zones. Moving to 3D, we build ROMs for 3D dynamic models of subduction and mantle flow, approximating viscosity, velocity and the second strain rate invariant in the Cascadia subduction zone. This study demonstrates that ROMs are effective tools that enable sensitivity analysis for high-dimensional 3D geodynamic models. Finally, we present a combined mesh morphing and model order reduction framework that allows us to quantify model sensitivity to geometric variability. We apply this method to meshes for 2D subduction zone thermal models, incorporating realistic slab interface curvature and depth uncertainties, and to 3D earthquake dynamic rupture models, varying fault dip angles across a 40° range. The mesh morphing method enables us to build ROMs that incorporate geometric variability, and we demonstrate this with ROMs built from dynamic rupture output that predict surface displacement and velocity time series for variable fault geometries. Together, these contributions provide a framework for robust sensitivity analysis of high-dimensional simulations for a range of geophysics problems.
■590 ▼aSchool code: 0033.
■650 4▼aGeophysics
■650 4▼aSoil sciences
■653 ▼aDynamic rupture
■653 ▼aReduced-order modeling
■653 ▼aSensitivity analysis
■653 ▼aSubduction zones
■690 ▼a0373
■690 ▼a0467
■690 ▼a0481
■71020▼aUniversity of California, San Diego▼bScripps Institution of Oceanography.
■7730 ▼tDissertations Abstracts International▼g87-06B.
■790 ▼a0033
■791 ▼aPh.D.
■792 ▼a2025
■793 ▼aEnglish
■85640▼uhttp://www.riss.kr/pdu/ddodLink.do?id=T17360078▼nKERIS▼z이 자료의 원문은 한국교육학술정보원에서 제공합니다.


