Unified framework for modelling progressive to catastrophic failure in fractured media

Projektbeschreibung

Fortschritte in Georisiken-Vorhersagewissenschaft

Katastrophale geologische Risiken vorherzusagen, bleibt eines der zentralen ungelösten Probleme der Geowissenschaften und im Geoengineering. Das Team des ERC-finanzierten Projekts FORECAST verfolgt das Ziel, die wissenschaftlichen Grundlagen der Georisikovorhersage zu verbessern. Dazu wird ein einheitlicher Rahmen zum Verständnis, zur Diagnose und Vorhersage katastrophaler Ereignisse im lokalen Maßstab, darunter Erdrutsche, Felsstürze, Gletscherabbrüche und Vulkanausbrüche, entwickelt. Mithilfe der Integration von Multiphysikmodellierung, statistischer Physik, nichtlinearer Dynamik und Hochleistungsrechnen untersucht das Projektteam, wie komplexe Wechselwirkungen innerhalb von gebrochenen Medien den Übergang von stabiler Verformung zu katastrophalen Störfällen auslösen. Das Team von FORECAST integriert physikbasierte Modellierungsansätze über mehrere Komplexitätsebenen hinweg mit Standortcharakterisierung und Beobachtungsdaten in einem hierarchischen Vorhersagerahmen, um die Naturgefahren-Vorhersagefähigkeiten der nächsten Generation voranzubringen.

Ziel

Climate change, characterised by rising temperatures, extreme rainfall, permafrost thaw, and sea level rise, poses great new challenges for mitigating extreme geohazards like mountain collapses, glacier breakoffs, and volcanic eruptions. Many of these extreme events are due to catastrophic failure in fractured media. There is a fundamental and urgent need to reliably predict these catastrophes and minimise their consequences. Extensive evidence indicates that geomaterials commonly incubate progressive failure, characterised by diffuse, gradual, and intermittent deformations, before culminating in catastrophic failure marked by localised, rapid, and runaway rupture. As of today, it remains unresolved why, how, and when progressive to catastrophic failure occurs in fractured media, inhibiting our ability to mitigate various extreme geohazards.

FORECAST proposes that catastrophic failure may be anticipated by analysing the interplay of three fundamental variables across scales in fractured media: heterogeneity (variability of material properties), nonlinearity (intricacy in physical processes), and susceptibility (vulnerability to external perturbations). FORECAST will develop a unified modelling framework to unravel this interplay and enable physics-based, path-dependent probabilistic forecast of the timing of catastrophic events. The framework will be tested on a large database of historical and ongoing geohazard events across seemingly distinct yet mechanistically interconnected contexts of landslides, rockbursts, glaciers, and volcanoes.

Three key questions will be addressed: (i) Why does progressive to catastrophic failure universally occur? (ii) How does failure evolve under variable and uncertain conditions? (iii) When will a catastrophic failure occur? FORECAST’s outcomes will have far-reaching implications for predicting catastrophic events in not only geological and engineered materials, but also social, economic, biological, and environmental systems.