The OpenQuake Engine started being developed in 2010 and was publicly introduced in 2014 as an open-source software for seismic hazard and risk modeling, aiming to provide a transparent, flexible, and globally accessible platform to the earthquake engineering and hazard science communities. Over the subsequent decade, extensive advancements have significantly expanded its capabilities and enhanced its adoption worldwide. This article comprehensively reviews these developments, detailing the new computational workflows and features implemented in the OpenQuake Engine with an emphasis on the risk component, improvements in its computational efficiency and scalability, and its growing global application across diverse geographical and thematic contexts. Major improvements to the OpenQuake Engine include the earthquake-triggered landslide and liquefaction modules, the infrastructure risk and network connectivity analysis module, the post-event loss amplification module, the financial loss module for insured and reinsured loss calculations, the classical and event-based probabilistic damage calculators, and the site-amplification module. Additional features include the option to condition ground motion fields on station data in scenario calculations, ability to connect ShakeMap outputs to the OpenQuake Engine’s damage and loss calculators, the conditional spectra calculator, vector-valued PSHA, and extension of the risk calculators to volcanic hazards. Performance improvements and enhancements in documentation have been pivotal in strengthening the software’s usability across a variety of computational platforms and user groups. A core contribution of this article lies in compiling and synthesizing over a hundred studies conducted using the OpenQuake Engine, thereby illustrating its versatile application at national, regional, urban, and site-specific scales. Moreover, examples of its use in earthquake insurance pricing and parametric catastrophe bond design are highlighted, demonstrating its practical relevance to risk management and financial resilience. Reflections on lessons learned regarding the importance of open-source practices, robust documentation, sustained user engagement, and interdisciplinary collaboration are discussed to inform future development and maintenance of scientific software with global applications.

This study investigates the evolution and current status of seismic design regulations in Argentina, Bolivia, Chile, Colombia, Ecuador, Peru, and Venezuela, using a pre-established methodology previously applied for Europe. It introduces a simplified methodology to estimate the proportion of buildings designed under four seismic code levels: no code, low code, moderate code, and high code. By analysing the progression of seismic design standards across South America, the study determines lateral force coefficients for a typical mid-rise reinforced concrete structure corresponding to each seismic code level. The findings reveal that approximately 20% of the total building stock, and 55% of reinforced concrete buildings, were constructed while regulations with some seismic provisions were followed. This research offers essential tools to enhance seismic risk assessment models and provides a dynamic framework for integrating new data, technological advancements, and local expertise into exposure modelling. Furthermore, it contributes to a global initiative led by the Global Earthquake Model (GEM) Foundation aimed at improving accessibility to information on seismic regulations and seismic hazard design demand maps.

Driven by rapid urbanization and heightened seismic risk concerns, efficient methods for developing regional seismic exposure assessments are advantageous. By leveraging deep learning and computer vision techniques, this study presents a novel approach for automating the identification of building typologies. The detailed building stock required for seismic exposure assessment has been traditionally achieved through time-consuming and costly in-person inspections. Recently, virtual inspections have emerged as a more efficient alternative, but they still require significant manual effort. This study proposes a methodology for automating the characterization of buildings, including details such as the number of stories, structural system, and construction period (pre-code or code), by implementing a convolutional neural network model that processes labeled images from Google Street View. A key innovation of this study is the integration of pre-processing techniques, including an object detector to isolate building façades and perspective correction using a keypoint model and homography transformation, enabling robust performance even with a small data set. This research advances prior methods by classifying individual stories rather than grouping them into broad taxonomic ranges, providing greater precision and applicability for seismic exposure modeling. The results show an 88% accuracy for structural system identification, a 78% accuracy for the number of stories, and a 69% accuracy for construction period determination. These characteristics are integrated into a probabilistic distribution model of building taxonomy that informs about their potential seismic vulnerability. The proposed procedures streamline the development of building stock and seismic exposure models, thus facilitating their use for seismic risk modeling at a regional scale.

The Global EarthquakE ScEnarios (GEESE) algorithm retrieves from a seismic hazard input model the ruptures matching a set of criteria (e.g., magnitude, location, focal mechanism). We applied the GEESE algorithm to create a publicly available database (version 1.0) of finite rupture models for past earthquakes which can be used for scenario seismic hazard and risk analysis applications. To this end, we selected earthquakes with a moment magnitude larger than 7.0 and hypocentral depth less than 200 km in the ISC-GEM catalogue (version 10.0) and retrieved the best matching ruptures from the seismic hazard models in the GEM Mosaic. The GEESE algorithm also automatically computes a set of ground-motion fields using each matched rupture, which are also provided in the database. The ability of the GEESE algorithm to test whether a Mosaic model can generate a rupture sufficiently representative of a queried event is a useful means of evaluating the Mosaic model's seismic source characterisation (SSC). Sufficiently matching ruptures are retrieved from the Global Mosaic for 90 percent of the tested ISC-GEM events. The GEESE algorithm can also be used in post-event response analysis to rapidly obtain an initial finite rupture when only minimal event information is initially available. A demonstration of these capabilities of the GEESE algorithm is provided using the 2023 Morocco earthquake, the 1994 Northridge earthquake, and the 2023 Kahramanmaras earthquake.

Several destructive earthquakes have occurred throughout the African continent over the past century. However, few comprehensive seismic risk models exist for the region. This study presents a probabilistic earthquake loss model for Africa, which is comprised of open model components and data sets that enable the calculation of a range of risk metrics useful for disaster risk management. Across the continent, Algeria faces the most significant predicted building damage, economic loss, population displacement, and fatality risks due to earthquakes. After Algeria, the order of highest risk countries depends on the considered risk metric, with countries such as Egypt, Morocco, and Uganda joining Algeria in the top three. When measured in relative terms, smaller countries that face disproportionate risks are highlighted, such as Djibouti, Burundi, Rwanda, and Malawi. These countries are exposed to moderate seismic hazard, but have limited evidence of earthquake-resistant design and construction practices that imply significant risk of damages in future earthquakes.

This study proposes a framework to evaluate probabilistic seismic risk of buildings in regions exposed to both crustal and subduction earthquakes. Ground motions from subduction earthquakes are typically longer in duration, causing higher damage to structure due to increased inelastic demand as compared to ground motions from crustal earthquakes having the same peak intensity. The increased vulnerability to structural damage from subduction earthquake ground motions needs to be accounted for in seismic loss assessment studies. This study investigates the same for India. The north and northeast of India are exposed to both crustal and continental subduction seismic sources, and Peninsular India is exposed to crustal seismic sources. Nonlinear analytical models are developed for a set of modern Indian code-compliant reinforced concrete special moment frame buildings located at 20 different sites in India. Incremental dynamic analysis (IDA) of building models using spectrally equivalent ground motions from crustal and subduction earthquakes is used for developing tectonic-region-type-specific (crustal and subduction) building vulnerability functions. The cumulative damage index is used as the engineering demand parameter to capture the increased inelastic demand from subduction earthquakes on buildings. For seismic risk assessment, the total seismic hazard at a site is separated into its contribution from crustal and subduction sources and combined with respective building vulnerability functions. The seismic risk of buildings is quantified by average annual loss ratio (AALR) through event-based probabilistic seismic risk analysis. For buildings located in high seismic zones of India, this study finds that AALR can be up to 40% higher on average as compared to studies not accounting for increased building vulnerability from subduction earthquakes.

This study presents the results of the seismic risk assessment for the city of Santiago de Cali (Colombia), a collaborative effort between the Mayor's Office of Santiago de Cali (including the Municipal Planning Office and the Disaster Risk Management Secretariat), the Colombian Geological Survey (SGC), the United States Geological Survey (USGS), EAFIT University and the GEM Foundation. The primary objective was to provide actionable insights for Disaster Risk Management (DRM) across its stages: awareness, risk reduction, and disaster response, and to enable the creation of a risk assessment specifically tailored to the city's needs, ensuring that its results can be effectively integrated into local risk management processes. Early engagement with stakeholders ensured that the models aligned with local needs, facilitating their integration into DRM policies. The seismic risk analysis utilizes the latest hazard model developed for Colombia, advanced non-linear site effects characterization, and a detailed building inventory with local expertise. For the first time, data-driven deterministic and probabilistic seismic risk estimates are presented for the city. The results indicate that destructive events could heavily impact the city, particularly events coming from the Dagua-Calima Fault or Nazca Plate subduction zone. Moreover, probabilistic outcomes show that human casualties and building damage disproportionally affect low-income areas, while almost 50 % of economic losses occur to middle-to high-income areas due to higher asset values. These findings highlight the importance of risk reduction strategies that address simultaneously the physical vulnerability and socio-economic disparities.

The Sendai Framework proposes recording disaster losses from hazard events between 2015 and 2030 to monitor the progress towards reduction targets to curtail risk worldwide. In the case of earthquakes, relying on losses over 15 years is unlikely to yield sufficient evidence to support risk mitigation strategies. In this study we propose a general methodology to monitor and forecast Sendai indicators. We apply our approach to explore trajectories of probabilistic indicators of mortality (A1) and economic loss (C1) in the form of custom Sendai Indicators for the Dominican Republic. Risk reduction targets, at national and subnational level, are established and tested along with two mitigation strategies: nationwide retrofitting campaign and stronger code-enforcement. The baseline projection indicates that earthquake risk is expected to increase at a national level, with indicators A1 and C1 increasing over 17% and 27%, respectively, in a period of 35 years. At that scale, code enforcement is more effective in meeting the targets of reduced mortality by 2030 and 2050. At the sub-national level, results depend on the risk drivers in each municipality. We provide two cases: in Santiago de los Caballeros, the baseline risk projection suggests that indicators A1 and C1 will increase by more than 18% and 26% by 2050, respectively, while in Distrito Nacional they decrease by almost 5% and 3%, respectively. In the former region, code-enforcement is the most effective measure to meet the risk reduction targets, while for the latter it was found to be the retrofit campaign.

This document contains frequently asked questions about public partnership opportunities with the GEM Foundation. It covers eligibility requirements for applicant organizations, guidelines for the application form, and the multi-stage selection process.

Occupancy patterns are known to strongly affect the number of people killed by earthquakes. Existing exposure models for Europe based on housing census do not account for the daily movement of the population between the place of residence (residential occupancy) and places of economic activity (non-residential occupancy), or the seasonal patterns due to tourism. This study presents a framework to upgrade exposure models from static to 'dynamic', i.e., allowing the input population to change in time and space based on daily and monthly population movement patterns. Open-source population data is used to disaggregate and rescale occupants inside residential, commercial and industrial buildings of 28 European countries, resulting in 24 occupancy categories: two times (i.e., day and night) x 12 months at 30 arc-seconds resolution. The static vs dynamic exposure models are compared using the number and distribution of fatalities resulting from loss calculations for a stochastic set of earthquakes generated from the European Seismic Hazard model (ESHM20). The results demonstrate that the spatiotemporal patterns of population can significantly impact earthquake mortality rates and should not be neglected in scenario loss assessment. The results also demonstrate that the worst occurrence time depends on both the distribution of indoor population between building occupancies and the earthquake rupture characteristics. The ability to capture population distribution during the day and night or seasonal changes (e.g., winter vs summer) is a feature that can advance the ongoing rapid damage/loss assessment services in Europe and consequently support emergency response planning.