Periodic Reporting for period 3 - FEASIBLe (Finding how Earthquakes And Storms Impact the Building of Landscapes)
Berichtszeitraum: 2022-06-01 bis 2023-11-30
Unraveling how tectonics, climate and surface processes act and interact to shape the Earth’s surface is one of the most challenging unresolved issue in Earth Sciences. The foundations of modern quantitative geomorphology have been built within the paradigm of steady-state landscapes responding to slow changes in climatic or tectonic conditions, mainly rainfall or uplift rate. Yet, recent results demonstrate that landscapes are rhythmed by (potentially extreme) storms and earthquakes. These perturbations catalyze geomorphological processes by triggering numerous landslides and lead to a prolonged and transient evolution of the landscape that dominate records of modern erosion. Moreover, these extreme events and their links through cascading processes represent threats to our societies which are poorly-constrained. The FEASIBLe project therefore calls for a complete re-assessment of the role of short-term climatic and tectonic perturbations in shaping mountain landscapes and for a paradigm shift from steady-state to constantly perturbed landscapes. The ambition of this project is to push forward our understanding of the short- to long-term dynamics of perturbed landscapes and in turn to unlock our ability to read landscapes in terms of earthquake and storm activity. To succeed in this endeavor, the FEASIBLe project will rely on the development of a new generation of landscape evolution model and of novel approaches to intimately monitor landscape heterogeneities and evolution in Taiwan, New-Zealand and Himalayas at high-resolution. The first work packages (WP1-2) will combine field-data analysis and numerical modelling to investigate landslide triggering and the post-perturbation sediment evacuation and landscape dynamics. We will then blend these elementary processes with a statistical description of climatic and tectonic perturbations in a new generation of landscape evolution model (WP3). This new model will be then applied to diagnose the geomorphological signature of fault “seismogenic” rheology (WP4) and to explore the role of post-glacial hot-moments of landscape dynamics on Quaternary landscape evolution (WP5).
Work performed and publications: The project has made progress on the three first work packages. Work packages 1 and 2 have advanced as planned. In WP2, we have benefitted from the experimental expertise of Laure Guerit to start some – not initially planned- experiments aiming at better understanding the role of fractures on grain size production, which represents one of the scientific challenges developed in the proposal. We have published a paper on the impact of sediment flux on river erosion and geometry (Baynes et al., 2020) and we are currently writing a paper on a new automatic approach to measure grain size from 3D data (Steer, Guerit et al., in prep.). We have also developed a new physical model to simulate the sediment transport and deposition of a multi-grain-size sediment population (Le Minor et al., in review) . In WP1, we have made some significant progress towards finding an optimal framework to model landslides and their triggering in landscape evolution models. This model was successfully tested against the first Lidar dataset of topographic changes induced by co-seismic landsliding, obtained in the Kaikoura region in New-Zealand (Bernard et al., 2021). We have published a paper on the impact of earthquakes on river profiles (Steer et al., 2019), one paper on landslide modelling in landscape evolution model (Campforts et al., 2020) and one paper where we use pre- and post-earthquake Lidar data in New Zealand to measure in 3D the topographic impact of co-seismic landslides (Bernard et al., 2021). We have also submitted a paper (Pelascini et al., in review) on the impact of atmospheric pressure change and rainfall on landslide triggering, that we are now applying to the numerous landslides triggered by the 2009 Morakot typhoon in Taiwan. In WP3, we have developed new statistical models to efficiently simulate landscape dynamics, including landslides and knickpoints, through several seismic cycles. We have also developed a new numerical model to efficiently simulate river hydrodynamics in landscape evolution models (Steer et al., in prep.).
Dissemination: Our results have been presented and discussed regularly in scientific conferences (e.g. EGU, AGU) and in media outlets. In particular, Philippe Steer has been a keynote speaker for the 2021 Steepest descent meeting, and he has contributed to a documentary on major catastrophic events that will be played on French TV in 2022. Philippe Steer has also initiated a new online series of weekly seminars, entitled "Landscapes Live", that is now co-organized by a team of 8 academics. Landscapes Live is dedicated to Geomorphological study and is freely open to anyone worldwide.
Dissemination: Our results have been presented and discussed regularly in scientific conferences (e.g. EGU, AGU) and in media outlets. In particular, Philippe Steer has been a keynote speaker for the 2021 Steepest descent meeting, and he has contributed to a documentary on major catastrophic events that will be played on French TV in 2022. Philippe Steer has also initiated a new online series of weekly seminars, entitled "Landscapes Live", that is now co-organized by a team of 8 academics. Landscapes Live is dedicated to Geomorphological study and is freely open to anyone worldwide.
Progress beyond the state of the art: Among our results, the paper by Bernard et al. (2021) probably represent the main scientific breakthrough we have achieved, with a large impact for natural hazard assessment. Indeed, this paper, that measured for the first time the direct topographic change induced by a population of co-seismic landslides, shows that even small landslides follow a power-law size distribution. We were surprised by these unexpected but rigorous outcome, which contradicts decades of research and hundreds of papers suggesting, based on landslides detected from 2D aerial optical images (and not 3D lidar topographic data), that landslide size distribution had a rollover for small sizes. This occurs probably due to a bias of measurement, similar to the well-known completeness magnitude in seismology, which makes the detection of small landslides more challenging when using 2D instead of 3D data. Our results also suggest that smallest landslides are more numerous than previously thought, which implies an increase in hazard and risk for populations and infrastructures.
Expected results: We expect to make significant progress on the modelling part of the project in the months and years to come. Our intention is to build a new open-access model to simulate the dynamics of large geomorphological perturbations. The first ingredients of this model are well underway, including the development of a fast river hydrodynamic model, a landslide model, a multi-grain-size sediment transport model and a statistical approach to simulate a series of perturbations (e.g. earthquakes). Technically, the challenge will be to merge all these different ingredients in a single landscape evolution model. Developing this new model will be fundamental to our following scientific actions planned in WP4 and WP5.
Expected results: We expect to make significant progress on the modelling part of the project in the months and years to come. Our intention is to build a new open-access model to simulate the dynamics of large geomorphological perturbations. The first ingredients of this model are well underway, including the development of a fast river hydrodynamic model, a landslide model, a multi-grain-size sediment transport model and a statistical approach to simulate a series of perturbations (e.g. earthquakes). Technically, the challenge will be to merge all these different ingredients in a single landscape evolution model. Developing this new model will be fundamental to our following scientific actions planned in WP4 and WP5.