OpenQuake Implementation of the Canterbury Seismic Hazard Model

Abstract

This article describes the release of the GNS Science Canterbury Seismic Hazard Model (CSHM), as implemented in the Global Earthquake Model’s OpenQuake software. Time‐varying models are implemented for the 50 yr time period between 2014 and 2064, as well as the 1 yr period from 1 September 2018 to 31 August 2019. Previous implementations have been confined to GNS in‐house software, and although source model input files have been made publicly available, this implementation improves the levels of visibility, documentation, and version control. Because of practical constraints in preparing a model for routine analysis, some corrections and changes to the previous implementations have been made. These constraints highlight issues for consideration when developing future hazard models, particularly the necessity of maintaining a balance between best‐practice science and practical model implementation. By implementing the CSHM in OpenQuake, the model is now in a form that allows users to obtain model outputs for engineering design, risk analyses, and prospective model testing.

Impact of Urbanization on Seismic Risk: A Study Based on Remote Sensing Data

“…This means that the results of the calculation presented in this paper are only theoretically reliable. Second, since the study area is relatively small, we perform calculations by setting a consistent PGA value rather than using the conventional seismic hazard model [80,81].…”

Challenges and opportunities in New Zealand seismic hazard and risk modeling using OpenQuake

“…As determined by Van Houtte and Abbott (2019a, 2019b), the CSHM was ultimately never fully implemented in the LFC due primarily to a combination of the CSHM’s complexity and the LFC’s shortcomings. That OQ is open source and enables more complexity in both source modeling and ground motion meant that, despite practical barriers, the CSHM could be implemented in full (Gerstenberger et al, 2014, 2016) in OQ without the additional coded workarounds required by the LFC.…”

“…That OQ is open source and enables more complexity in both source modeling and ground motion meant that, despite practical barriers, the CSHM could be implemented in full (Gerstenberger et al, 2014, 2016) in OQ without the additional coded workarounds required by the LFC. This includes the use of logic trees to accommodate for the use of multiple GMPEs, whose specific modifications were incorporated into the OQ engine, and the use of logic trees to enable more streamlined use of the CSHM’s nine alternative source models that were also adapted for OQ (Van Houtte and Abbott, 2019a, 2019b). However, the KSHM was implemented directly into OQ, bypassing many of the challenges experienced in benchmarking the NSHM and implementing the CSHM, and making use of OQ’s more extensive capabilities.…”

“…OQ-produced hazard results for the NSHM, CSHM (Van Houtte and Abbott, 2019a, 2019b), and KSHM are shown in Figure 4. They demonstrate clearly elevated regional hazard relative to the surrounding NSHM.…”

Forensic PSHA: Benchmarking Canada’s Fifth Generation seismic hazard model using the OpenQuake-engine

“…Finally, the adoption of open-source platforms can enable end-users to more readily ingest and adapt published hazard input models according to their specific use cases (e.g. Allen et al, 2018; Van Houtte and Abbott, 2019). These benefits must, however, be balanced against the cost of investment of resources for an organization to implement contributions, whether support is sufficient (be it from the original software developers or the active user community), and ultimately whether the software has the likelihood of longevity in terms of active development and maintenance.…”