UCVM Workshop

UCVM is an software system for hosting and querying 3D velocity models. Currently, multiple California velocity models are supported in UCVM.

To support development of UCVM, we are holding "virtual" workshops, that means teleconferences, to discuss and plan both scientific and technical issues regarding its development.

Workshop July 2011

UCVM Software Design - Framework for hosting multiple CVMs - Standardized query interface - How UCVM supports queries for non-standard models (e.g. How does UCVM handle elevation query using CVM-S4 (no elevation model)

Selected standardized statewide properties including: - Topography - Bathymetry - Statewide Geotechnical layer from Vs30 and Wald Vs30 topography relationship - Georeference standardization

Selected UCM Coverage Region: - Discuss portability of UCVM other regions including Central US

Currently Supported Models - CVM-S4 - CVM-SI - CVM-H 11.2 - CVM-H 6.3 - CenCal USGS - Graves Northern Cal - Lin Thurber statewide

Requirements for Hosted Models - query interface - projection definition

Options for storing hosted CVMS - GTL etree - CVM-S4 objects - CVM-H gocad - Lin-Thurber tables - Po Chen CVM-SI mesh

Support for HPC creation of Large Meshes by UCVM - parallel query capabilities - large mesh making capabilities

ETREE support - Description of etree import and export by UCVM

CVM Differencing capabilities - Example differences between CVM-S4 and CVM-SI - Proposed differencing tools - Other CVM differences of interest

CVM Evaluation System Design - Evaluation using point sources - Evaluation using large historical events - Goodness of fit measures in use - Proposed GOF measures to be used

Evaluation results for 3 events - Evaluate sensitivity of current evaluation system using 3 events and 3 CVM’s - Chino Hills,Gilroy, Oakland and CVM-H, CVM-S4, CenCal

Workshop May 2011

California State-wide Community Velocity Model (CVM) Development: a State-wide 3D Velocity Model for California

Workshop Date: Tuesday 10 May 2011 Workshop Time: 9:30am – 11:00am (90 minutes) Agenda Version: 8 May 2011

1. We plan to hold two CVM development workshops via teleconference calls prior to the next SCEC Annual Meeting (which will be held Sept 11-14, 2011). 1.1. First CVM workshop to discuss and establish CVM development goals (10 May 2011) 1.2. Second CVM workshop to discuss CVM development plans (week of June 13, 2011)

2. Several trends are pushing SCEC towards development of state-wide models: 2.1. State-wide Earthquake Rupture Forecast (UCERF) development 2.2. CISN and ANSS 2.3. Operational Earthquake Forecasting 2.4. Simulation for large California regions 2.5. SCEC4 goals 2.6. Statewide wave propagation-based PSHA calculations (CyberShake)

3. SCEC context supporting development of state-wide CVM 3.1. SCEC’s long standing scientific goal of compatible, integrated, computational structural models to support seismic hazard computational models including fault stress models, dynamic rupture simulations, wave propagation models, earthquake simulators: 3.1.1.1. CVM 3.1.1.2. CFM 3.1.1.3. Block Model 3.1.1.4. Crustal Motion Model 3.2. UCERF2.0 development 3.3. UCERF3.0 development 3.4. Release of multiple southern California CVM’s (CVM-S4 and CVM-H 11.2) 3.5. Introduction of alternative Geotechnical layer models-based on Vs30

4. Review of CVM User Communities to be considered: 4.1. High frequency wave propagation simulations (>1Hz) 4.2. Low frequency wave propagation simulations (< 1Hz) 4.3. Dynamic rupture modeling 4.4. Seismic hazard PSHA calculations 4.5. Others to be considered?

5. Discussion of Capabilities of Proposed state-wide CVM 5.1. Region to be covered 5.2. What features will it have (GTL, bathymetry, topography) 5.3. What users will is support 5.4. Resolution 5.5. Expression of uncertainty 5.6. Arbitrary precision 5.7. Expression of maximum useful frequency or resolution

6. Model construction, implementation, and delivery 6.1. New model development 6.2. Combining or integrating best of existing models 6.3. Management of alternative models 6.4. Blending of existing models 6.5. Delivery of model to end-users 6.6. Delivery of model to model-developers

7. Evaluation of Candidate and Alternative Models 7.1. Proposed use of wave propagation simulations and comparison of seismograms to observations to evaluate alternative models 7.2. Description of existing SCEC CVM evaluation system (developed for CVM-H,CVM-S) 7.3. Additional capabilities needed in an automated testing system which can repeat the same simulation and same evaluation tests with alternative models 7.4. Selection of goodness of fit metrics 7.5. Selection of goodness reference earthquakes distributed around the state and associated observational datasets 7.6. Level of support by SCEC for alternative models

8. Review of existing models 8.1. CVM-S4 (Magistrale) 8.2. CVM-S4 3D inversion update (Chen) 8.3. CVM-H 11.2 (Shaw) 8.4. CVM-H-based Central California update (Tape) 8.5. USGS Central California Model (Aagaard) 8.6. Lin-Thurber state-wide model (Lin) 8.7. Hauksson model (Hauksson) 8.8. Graves Mendicino Model (Graves) 8.9. Thurber Northern California Model (Thurber) 8.10. Reno Nevada Model (Magistrale) 8.11. Las Vegas Nevada Model (Magistrale)

9. External Groups with CVM Development Tools, Interests, Capabilities 9.1. IRIS 9.2. USGS 9.3. Others?

10. Technical Development Approach 10.1. Development of CVM software: 10.2. UCVM software is intended to support statewide-CVM development. 10.3. Supports standardized interface that can access multiple underlying models 10.3.1. CVM-S4 10.3.2. CVM-H 11.2 10.3.3. USGS Central California Model 10.3.4. Lin-Thurber 10.3.5. 1D background models 10.4. Supports state-wide topography, bathymetry models 10.5. Supports application of state-wide geotechnical layer based on Wills/Wald Vs30 10.6. Supports creation and delivery of etree models 10.7. Supports query by depth and query by elevation 10.8. Supports creation of large meshes

10. Tools required for Model Development 10.1. Automated CVM evaluation capabilities using wave propagation simulations 10.2. Mesh construction 10.3. Combining alternative models (e.g. Lin-Thurber with CVM-H 11.2) 10.4. Mesh slices and profiles 10.5. Model volume visualization 10.6. Migration between existing model formats (USGS e-tree, Gocad voxels, CVM-S objects, Lin-Thurber tables)

11. Software development activities and priorities 11.1. Discussion of current UCVM capabilities 11.2. Discussion of software development priorities