Search

The impact of glaciers on the Quaternary evolution of mountainous landscapes remains controversial. Although in situ or bedrock low-temperature thermochronology offers insights on past rock exhumation and landscape erosion, the method also suffers from potential biases due to the difficulty of sampling bedrock buried under glaciers. Detrital thermochronology attempts to overcome this issue by sampling sediments at e.g. the catchment outlet, a component of which may originate from beneath the ice. However, detrital age distributions not only reflect the catchment exhumation, but also spatially variable patterns and rates of surface erosion and sediment transport. In this study, we use a new version of a glacial landscape evolution model, iSOSIA, to address the effect of erosion and sediment transport by ice on the form of synthetic detrital age distributions. Sediments are tracked as Lagrangian particles formed by bedrock erosion, and their transport is restricted to ice or hillslope processes, neglecting subglacial hydrology, until they are deposited. We base our model on the Tiedemann Glacier (British Columbia, Canada), which has simple morphological characteristics, such as a linear form and no connectivity to large tributary glaciers. Synthetic detrital age distributions are generated by specifying an erosion history, then sampling sediment particles at the frontal moraine of the modelled glacier. Results show that sediment sources, reflecting different processes such as glacier and hillslope erosion, can have distinct bedrock age distribution signatures, and estimating such distributions should help to identify predominant sources in the sampling site. However, discrepancies between the detrital and bedrock age distributions occur due to (i) the selective storage of a large proportion of sediments in small tributary glaciers and in lateral moraines, (ii) the large range of particle transport times due to varying transport lengths and strong variability of glacier ice velocity, (iii) the heterogeneous pattern of erosion, and (iv) the advective nature of glacier sediment transport along ice streamlines. This last factor leads to a poor lateral mixing of particle detrital signatures inside the frontal moraine, and then local sampling of the frontal moraine is likely to reflect local sources upstream. Therefore, sampling randomly across the moraine is preferred for a more representative view of the catchment age distribution. Finally, systematic comparisons between synthetic (U-Th)/He and fission track detrital ages, with different bedrock age-elevation profiles and different relative age uncertainties, show that the nature of the age-elevation relationship and age uncertainties largely control the ability to track sediment sources in the detrital record. However, depending on the erosion pattern spatially, qualitative first-order information may still be extracted from a thermochronological system with high uncertainties (>30 %). Overall, our results demonstrate that detrital age distributions in glaciated catchments are strongly impacted not only by erosion and exhumation but also by sediment transport processes and their spatial variability. However, when combined with bedrock age distributions, detrital thermochronology offers a novel means to constrain the transport pattern and time of sediment particles.

International audience; The Taiwan mountain range stands as one of the most active regions on Earth. With an overall shortening rate of ~40 mm/yr and an average erosion rate of ~4 mm/yr, this mountain range appears ideal to better understand the interactions between tectonics and surface processes, and how these shape active landscapes. Here we explore the geomorphic and sedimentary record of active deformation within the Southwestern Foothills of Taiwan, and we quantify from there the kinematics of active faults. In particular, we investigate the downstream portion of the meandering Tsengwen river - one of the largest rivers of this region - where we identify and correlate remnants of 7 terrace levels, progressively abandoned over the last ~5 kyr. The incision of these terraces is interpreted as being controlled to the first-order by folding and uplift related to underlying active faults. The evolution of the river is reconstructed from correlated terrace remnants, and our results indicate that the overall river sinuosity and gradient did not vary significantly during the past ~5 kyr in response to tectonics. Incremental tectonic uplift is retrieved from terrace incision corrected for sedimentation at the mountain front, and is used to derive the incremental shortening since terrace abandonment. Downstream, within the Coastal Plain, the Tsengwen river reaches its base level and aggrades. Sedimentary facies within boreholes of the Coastal Plain record vertical displacements relative to sea level, spatially consistent with potential blind active faults. When corrected for eustatic variations, these data allow for quantifying tectonic uplift rates within the Coastal Plain over the last ~20 kyr. Taken altogether, our quantitative analysis of the Tsengwen river record, from terrace incision to dowsntream aggradation, reveals that the most frontal active faults absorb a shortening rate of at least ~35 mm/yr, that is most of - if not all - the shortening rate to the absorbed across the whole mountain range.

International audience; The role of variable sediment supply is often overlooked in modelling studies of landscape evolution, despite sediment playing a key role in the physical processes that drive erosion and sedimentation in river channels. Under a changing climate, or in active tectonic settings, landscapes face pressures from variable sediment supply in the future, with the resulting impact on channel dynamics currently uncertain. Here, we present a combined field and physical modelling study that demonstrates the importance of feedbacks between sediment supply and channel/valley geometry in bedrock, gravel, and mixed gravel-bedrock river systems. In the Rangitikei river catchment, New Zealand, bedrock channels receiving a supply of coarse, hard, sediment are systematically wider than channels with no sediment input for a given discharge. Box flume experiments (0.8 x 0.3 m) exploring this dependency further demonstrate that multiple configurations of bedrock and mixed bedrock-gravel channel geometry can exist for a given discharge, solely due to input sediment supply. Larger flume experiments (12 x 2.4 m) examine the dynamics of analogue gravel-bed river systems, including a setup where the channel was forced through a narrow constriction relative to its unconstrained width, replicating the bedrock valley geometry common to many gravel-bed river systems that transport large volumes of sediment in New Zealand. Experiments were run with a constant sediment supply until they reached dynamic equilibrium, before the sediment supply was increased at the channel inlet, representing scenarios of heightened sediment availability from upland areas driven by a shift to a more extreme climate. The impact of a valley constriction on the channel dynamics was identified through the comparison of the response to increased sediment supply for channels with and without the constriction. These results will help to develop the process-based understanding of channel dynamics under variable forcing regimes, which can be used to aid the development of numerical models and a host of applications including landscape evolution or flood risk.

International audience; Landscapes in Greenland have been deeply impacted by glacial erosion since the major cooling event at theEocene-Oligocene boundary and the initiation of the Greenland ice sheet between 5 and 30 Ma. Glacial erosioncreated deep and wide fjords including the largest in the world, the Scoresby Sund fjord, with a width greater than40 km. As fjords are unequivocally associated to glacial erosion, they represent a unique archive to assess theco-evolution of ice dynamics and landscapes during the Cenozoic. More specifically, the thermal history of rocksassessed from low-temperature thermochronology offer valuable insights to infer the long-term erosion history ofGreenland landscapes and the spatial distribution of ice over time. In the last decade, detrital thermochronologyhas been used to deduce spatial patterns of erosion within a catchment by considering just age probabilitydistributions functions (PDF). Recently, we have developed a new inverse approach to infer thermal historiesdirectly from detrital age distributions, without the need for in situ or vertical profile bedrock data.As this approach has shown promising results, we built the concept into the 3D landscape evolution model iSOSIA.Within the model we track sediment-particles resulting from erosion through multiple glaciation events, leadingto the formation of a major fjord comparable to those observed in Greenland. Each particle has an associatedthermochronological age, reflecting both the age distribution on the pre-glacial landscape and any modifications inresponse to glacial erosion. This information allows us to predict age distributions in a detrital sample anywherein the catchment. From these distributions we are then able to infer thermal histories, and to model the patternof erosion using an inverse approach. We present results of this modelling approach in a case study simulatingfjord development in East Greenland. The results highlight the importance of the strategy of detrital sampling forconstraining erosion history of fjords and the impact of Cenozoic glaciations on landscape evolution.

International audience; In active mountain ranges, earthquakes can trigger widespread landslides that mobilise large volumes of sediment. This sudden sediment delivery can cause downstream changes in river geometry and transport capacity that affect the river efficiency to export this sediment out of the epicentral area. The propagation of sediment from landslide deposits through the fluvial network has implications for the management of hazards downstream and for landscape dynamics over time scales of the seismic cycle. However, a full understanding of the processes and time scales associated with post-seismic landslide removal by bedload transport is still lacking. Here, we propose two numerical approaches to assess the evacuation time of landslide material at different spatial and time scales. First, we explore the river morphodynamic response to a landslide occurrence at the reach-scale using a 2D modelling approach. We use a simplified bedrock channel to study the influence of the volume of sediment pulse (VS) and channel transport capacity (QT) on the export time of landslides. Two regimes are identified: (i) the export time is linearly related to VS/QT when the sediment pulse does not affect significantly river hydrodynamics (low VS/QT) and (ii) the export time is a non-linear function of VS/QT when the pulse undergoes significant morphodynamic modifications during its evacuation (high VS/QT). In the latter case, active narrowing of the river within the landslide deposit is responsible for a significant increase in the transport capacity, resulting in faster evacuation times than predicted by theory. Secondly, we propose a statistical approach to quantify evacuation times of clusters of earthquake-triggered landslides at the scale of mountain ranges. Our approach combines an empirical description of co-seismic landslide clusters with the physical processes involved during the post-seismic phase and focuses on hypothetical Mw 8 earthquakes occurring on the Alpine Fault, New Zealand. In particular, we explore the role of landslide connectivity on the modulation of post-seismic sediment fluxes. We also show how this approach could shed light on sediment fluxes from active mountain ranges over longer timescales (i.e. several seismic cycles) and its potential to be coupled to a 2D landscape evolution model.

International audience; The Pyrenean Range results from the collision between Eurasia and Iberia plates that started during the Santonian. The Pyrenean orogeny is considered to end in Early Miocene times, since from that period onwards, Iberia shows no more differential motion with respect to the rest of Europe and becomes part of the Eurasian plate. This coincided with the slowing down of African-Eurasian convergence. Nevertheless, the Pyrenean relief still displays an unexpected strong relief and a combination of various geological methods highlights the occurrence of local Upper Miocene to present-day uplifts in the Pyrenees (low temperature thermochronological dating, sedimentology, karstology, palynology, palaeoaltimetry). While it has already been studied extensively in the Eastern Pyrenees, fewer works refer to the Western Pyrenean uplift. Depending on the method considered, the timing and the vertical amplitude of this uplift differs.We are initiating a geomorphological study of the incisions in the Western Pyrenees in order to constrain this post-shortening uplift. We focused on the Aspe Valley, where Miocene highly elevated – low relief erosion paleosurfaces observed as relicts on top of the interfluves have been strongly dissected. This dissection was achieved in several steps, including a low erosion episode, followed by an acceleration of erosion which could be attributed to the Late Miocene to actual Western Pyrenean uplift event. The sharp incisions are highlighted by erosion triangular facets which cannot be attributed to glacial erosion as they occur beyond the last glacial deposits. We attempt to reconstitute the landscape prior to this late incision in order to quantify the subsequent erosion and to try to link it with foreland deposits.Evidence of a Late Miocene to present-day uplift is well described in the Eastern Pyrenees and at a larger scale, in the Central Massif, the peri-alpine area and in northern and central Iberia thus suggesting an uplift at a lithospheric scale. However, discrepancies in the uplift rates indicated by already published data are found at a kilometric scale in the Western Pyrenees, which leads us to consider a possible contribution of both local and regional processes to explain the recent incisions in the Aspe Valley.This work is funded and carried out in the framework of the BRGM-TOTAL project Source-to- Sink.

International audience; Quantifying continental erosion brings valuable constraints on tectonics, climate or human activities, which all havean active role in surface processes. This quantification is classically done through in-situ measurements of riversediment discharge but often suffers from spatio-temporal scarcity. Here we show that temporal gravimetry, that is,the integrative measure of mass changes through time, is an appropriate tool for such a quantification, at both localand global scales, provided geometric constraints on the redistributed sediment are available. At the local scale,we combined repeated terrestrial gravity and drone photogrammetry surveys, once a year in 2015, 2016 and 2017over a 1-km2 area in central Taiwan featuring both a large river and a slow landslide. The good agreement betweengravity and terrain changes allowed us to retrieve the mass of sediment transiting locally in this area. At the globalscale, we take advantage of satellite gravimetry (GRACE), which has monitored gravity changes worldwide for 15years. By coupling sedimentations zones estimates derived from sea currents model ECCO2 and GRACE satellitedata, we retrieved sediment fluxes consistent with in situ data offshore rivers with the largest sediment loads. Atthe light of these experimental studies, the ongoing efforts to improve both gravimeters and geomorphometric dataopen new perspectives in the field of quantitative geomorphology. In particular, key modern challenges such asclosing the budget between erosion and sedimentation or monitoring sand mining, a rising societal concern, couldbe efficiently tackled by this interdisciplinary approach.

International audience; Landscape evolution models generally solve for the time variation of elevation due to erosional processes on afixed grid. This Eulerian description offers a tractable approach, despite leading to unnecessary dependency toslope, and in turn to numerical diffusion, when solving for the specific stream power incision model, e.g. with aslope exponent equal to 1. Indeed, under this condition, the stream power model reduces to a linear kinematicwave equation with a wave celerity independent of the local slope (Whitham, 1974; Whipple & Tucker, 1999).Using a Lagrangian description and solving for the horizontal advection of the topography due to erosion enablesto develop a simple finite difference numerical scheme fully independent of local slope. This scheme presentsnumerous benefits over Eulerian descriptions, including the absence of numerical diffusion, unconditional stabilityand the possibility to track geomorphological features and slope instabilities such as knickpoints. We exploit thelatter to track co-seismic knickpoints in simulations where uplift history is computed from a statistical modelof earthquakes (Turcotte et al., 2007; Leonard 2010). This model results in spatial and temporal distributionsof earthquakes that are consistent with Gutenberg-Richter frequency-magnitude scaling, a modified Båth’s law,Omori’s law in space and time (Turcotte et al., 2007) and earthquake dimension scaling laws (Leonard 2010).Motivated by recent results from sandbox experiments (Baynes, 2016), we then introduce a dependency ofknickpoint celerity to knickpoint height, that is initially here equals to co-seismic displacement. This modelauthorizes the merging of successive and colliding knickpoints due to different celerities. We use this model toinvestigate the role of knickpoint height and temporal distribution and of the dependency of their celerity to heighton the shape of river profiles. This model leads to river profiles with concavities deviating from predicted ones andto particular morphologies, including river profile characterized by a succession of knickzones separated by flatzones. We then discuss these implications in the light of natural river profiles.

International audience; It is generally observed that the dimensions of large bedrock landslides follow power-law scaling relationships.In particular, the non-cumulative frequency distribution (PDF) of bedrock landslide area is well characterized bya negative power-law above a critical size, with an exponent 2.4. However, the respective role of bedrock mechanicalproperties, landscape shape and triggering mechanisms on the scaling properties of landslide dimensionsare still poorly understood. Yet, unravelling the factors that control this distribution is required to better estimatethe total volume of landslides triggered by large earthquakes or storms. To tackle this issue, we develop a simpleprobabilistic 1D approach to compute the PDF of rupture depths in a given landscape. The model is applied torandomly sampled points along hillslopes of studied digital elevation models. At each point location, the modeldetermines the range of depth and angle leading to unstable rupture planes, by applying a simple Mohr-Coulombrupture criterion only to the rupture planes that intersect downhill surface topography. This model therefore accountsfor both rock mechanical properties, friction and cohesion, and landscape shape. We show that this modelleads to realistic landslide depth distribution, with a power-law arising when the number of samples is high enough.The modeled PDF of landslide size obtained for several landscapes match the ones from earthquakes-driven landslidescatalogues for the same landscape. In turn, this allows us to invert landslide effective mechanical parameters,friction and cohesion, associated to those specific events, including Chi-Chi,Wenchuan, Niigata and Gorkha earthquakes.The cohesion and friction ranges (25-35 degrees and 5-20 kPa) are in good agreement with previouslyinverted values. Our results demonstrate that reduced complexity mechanics is efficient to model the distributionof unstable depths. It also shows the importance of landscape variability in landslide size distribution and in thetotal volume of sediments produced by one triggering event.

International audience; Worldwide rivers annually export about 19 Gigatons of sediments to the ocean that mostly accumulate in the coastalzones and on the continental shelves. This sediment discharge testifies of the intensity of continental erosionand records changes in climate, tectonics and human activity. However, natural and instrumental uncertaintiesinherent to the in-situ measurements of sediment discharge prevent from conclusive estimates to better understandthese linkages. Here we develop a new method, using the Gravity Recovery and Climate Experiment (GRACE)satellite data, to infer mass-integrative estimates of sediment discharge of large rivers to the ocean. GRACE satelliteprovides global gravity time series that have proven useful for quantifying mass transport, including continentalwater redistribution at the Earth surface (ice sheets and glaciers melting, groundwater storage variations) but hasbeen seldom used for monitoring sediment mass transfers so far. Here we pair the analysis of regularized GRACEsolutions at high spatial resolution corrected from all known contributions (hydrology, ocean, atmosphere) to aparticle tracking model that predicts the location of the sediment sinks for 13 rivers with the highest sedimentsloads in the world. We find that the resulting GRACE-derived sediment discharges off the mouth of the Amazon,Ganges-Brahmaputra, Changjiang (Yangtze), Indus, Magdalena, Godavari and Mekong rivers are consistent within-situ measurements. Our results suggest that the lack of time continuity and of global coverage in terrestrialsediment discharge measurements could be reduced by using GRACE, which provides global and continuous datasince 2002. GRACE solutions are regularly improved and new satellite gravity missions are being prepared hencemaking our approach even more relevant in a near future. The accumulation of sediments over time will keepincreasing the signal to noise ratio of the gravity time series, which will improve the precision of the GRACE-derived sediment discharges values.

International audience; Extreme rainfall events can trigger numerous landslides in mountainous areas and a prolonged increase of river sediment load. The resulting mass transfer at the Earth surface in turn induces stress changes at depth, which could be sufficient to trigger shallow earthquakes. The 2009 Morakot typhoon represents a good case study as it delivered 3 m of precipitation in 3 days and caused some of the most intense erosion ever recorded. Analysis of seismicity time-series before and after the Morakot typhoon reveals a systematic increase of shallow (i.e. 0-15 km of depth) earthquake frequency in the vicinity of the areas displaying a high spatial density of landslides. This step-like increase in frequency lasts for at least 2-3 years and does not follow an Omori-type aftershock sequence. Rather, it is associated to a step change of the Gutenberg-Richter b-value of the earthquake catalog. Both changes occurred in mountainous areas of southwest Taiwan, where typhoon Morakot caused extensive landsliding. These spatial and temporal correlations strongly suggest a causal relationship between the Morakot-triggered landslides and the increase of earthquake frequency and their associated b-value. We propose that the progressive removal of landslide materials from the steep mountain landscape by river sediment transport acts as an approximately constant increase of the stress rate with respect to pre-typhoon conditions, and that this in turn causes a step-wise increase in earthquake frequency. To test this hypothesis, we investigate the response of a rate-and-state fault to stress changes using a 2-D continuum elasto-dynamic model. Consistent with the results of Ader et al. (2013), our preliminary results show a step-like increase of earthquake frequency in response to a step-like decrease of the fault normal stress. We also investigate the sensitivity of the amplitude and time-scale of the earthquake frequency increase to the amplitude of the normal stress change and to rheological parameters. Our study offers new insights on the potential influence of extreme erosional events on the short-time scale dynamics of faults and earthquakes.

International audience; Extreme rainfall events can trigger numerous landslides in mountainous areas and a prolonged increase of riversediment load. The resulting mass transfer at the Earth surface in turn induces stress changes at depth, whichcould be sufficient to trigger shallow earthquakes. The 2009 Morakot typhoon represents a good case study as itdelivered 3 m of precipitation in 3 days and caused some of the most intense erosion ever recorded. Analysis ofseismicity time-series before and after the Morakot typhoon reveals a systematic increase of shallow (i.e. 0-15 kmof depth) earthquake frequency in the vicinity of the areas displaying a high spatial density of landslides. Thisstep-like increase in frequency lasts for at least 2-3 years and does not follow an Omori-type aftershock sequence.Rather, it is associated to a step change of the Gutenberg-Richter b-value of the earthquake catalog. Both changesoccurred in mountainous areas of southwest Taiwan, where typhoon Morakot caused extensive landsliding. Thesespatial and temporal correlations strongly suggest a causal relationship between the Morakot-triggered landslidesand the increase of earthquake frequency and their associated b-value. We propose that the progressive removalof landslide materials from the steep mountain landscape by river sediment transport acts as an approximatelyconstant increase of the stress rate with respect to pre-typhoon conditions, and that this in turn causes a step-wiseincrease in earthquake frequency. To test this hypothesis, we investigate the response of a rate-and-state fault tostress changes using a 2-D continuum elasto-dynamic model. Consistent with the results of Ader et al. (2013), ourpreliminary results show a step-like increase of earthquake frequency in response to a step-like decrease of thefault normal stress. We also investigate the sensitivity of the amplitude and time-scale of the earthquake frequencyincrease to the amplitude of the normal stress change and to rheological parameters. Our study offers new insightson the potential influence of extreme erosional events on the short-time scale dynamics of faults and earthquakes

International audience; Mass wasting caused by large magnitude earthquakes choke mountain rivers with several cubic kilometers of sediment.The timescale and mechanisms by which rivers evacuate the coarse fraction of small to gigantic landslidedeposits are poorly known, but are critical to predict post-seismic hydro-sedimentary hazards, interpret the signatureof earthquakes in sedimentary archives and decipher the coupling between erosion and tectonics. Here, weuse a new 2D hydro-sedimentary evolution model to demonstrate that river self-organization into a narrower alluvialchannel overlying the bedrock valley dramatically increases sediment transport capacity of coarse sedimentsand reduces export time of gigantic landslides by orders of magnitude compared to existing theory. Predicted exporttimes obey a universal non-linear relationship function of landslide volume and pre-landslide valley transportcapacity. Dynamic alluvial channel narrowing is therefore a key, previously unrecognized, mechanism by whichmountain rivers rapidly digest extreme events and maintain their capacity to incise uplifted rocks. Upscaling theseresults to realistic populations of landslides show that removing half of the total sediment volume introduced bylarge earthquakes in the fluvial network would typically last 5 to 25 years in various tectonically active mountainbelts, with little impact of topography and climate. If several studies indicate a strong dependency of total landslidevolume to earthquake magnitude, our study show that the sediment export time of a landslide population isnot strongly impacted by earthquake magnitude or by the total volume of the landslide population. Building onthese new findings, we then investigate the dynamics of mountainous landscapes submitted to a series of earthquakes,following either a Gutenberg-Richter distribution or a single large magnitude event. We infer the temporaland spatial evolution of the number of active landslide deposits, of the sediment load along the fluvial networkand of the exported sediment flux throughout several seismic cycles. These results highlight how landscapes andsediment fluxes respond on longer time scales to a succession of earthquakes able to trigger landslides

International audience; In active mountain ranges, large earthquakes (Mw > 5-6) trigger numerous landslides that impact river dynamics. These landslides bring local and sudden sediment deposits which are eroded and transported along the river network, causing downstream changes in river geometry, transport capacity and erosion efficiency. The progressive removal of landslide materials has implications for downstream hazards management and for landscape dynamics at the timescale of the seismic cycle. Although the export time of suspended sediments from landslides triggered by large-magnitude earthquakes has been extensively studied, the processes and time scales associated to bedload transport remains poorly studied. Here, we study the sediment export of large landslides with the 2D morphodynamic model, Eros. This model combines: (i) an hydrodynamic model, (ii) a sediment transport and deposition model and (iii) a lateral erosion model. Eros is particularly well suited for this issue as it accounts for the complex retro-actions between sediment transport and fluvial geometry for rivers submitted to external forcings such as abrupt sediment supply increase. Using a simplified synthetic topography we systematically study the influence of pulse volume (Vs) and channel transport capacity (QT) on the export time of landslides. The range of simulated river behavior includes landslide vertical incision, its subsequent removal by lateral erosion and the river morphology modifications induced by downstream sediment propagation. The morphodynamic adaptation of the river increases its transport capacity along the channel and tends to accelerate the landslide evacuation. Our results highlight two regimes: (i) the export time is linearly related to Vs/QT when the sediment pulse introduced in the river does not affect significantly the river hydrodynamic (low Vs/QT) and (ii) the export time is a non-linear function of Vs/QT when the pulse undergoes significant morphodynamic modifications during its evacuation (high Vs/QT). By combining our newly derived export time functions with the frequency-magnitude of earthquake intensity and the induced sediment production, we investigate the sediment export of several plausible earthquake scenarii in different mountain ranges (New Zealand, Taiwan, Nepal).

International audience; Extreme rainfall events can trigger numerous landslides in mountainous areas and a prolonged increase of river sediment load. The resulting mass transfer at the Earth surface in turn induces stress changes at depth, which could be sufficient to trigger shallow earthquakes. The 2009 Morakot typhoon represents a good case study as it delivered 3 m of precipitation in 3 days and caused some of the most intense erosion ever recorded. Analysis of seismicity time-series before and after the Morakot typhoon reveals a systematic increase of shallow (i.e. 0-15 km of depth) earthquake frequency in the vicinity of the areas displaying a high spatial density of landslides. This step-like increase in frequency lasts for at least 2-3 years and does not follow an Omori-type aftershock sequence. Rather, it is associated to a step change of the Gutenberg-Richter b-value of the earthquake catalog. Both changes occurred in mountainous areas of southwest Taiwan, where typhoon Morakot caused extensive landsliding. These spatial and temporal correlations strongly suggest a causal relationship between the Morakot-triggered landslides and the increase of earthquake frequency and their associated b-value. We propose that the progressive removal of landslide materials from the steep mountain landscape by river sediment transport acts as an approximately constant increase of the stress rate with respect to pre-typhoon conditions, and that this in turn causes a step-wise increase in earthquake frequency. To test this hypothesis, we investigate the response of a rate-and-state fault to stress changes using a 2-D continuum elasto-dynamic model. Consistent with the results of Ader et al. (2013), our preliminary results show a step-like increase of earthquake frequency in response to a step-like decrease of the fault normal stress. We also investigate the sensitivity of the amplitude and time-scale of the earthquake frequency increase to the amplitude of the normal stress change and to rheological parameters. Our study offers new insights on the potential influence of extreme erosional events on the short-time scale dynamics of faults and earthquakes.

International audience; Coseismic landsliding is an important contributor to the long-term erosion of mountain belts. The scaling between earthquake magnitude and the volume of eroded material has been known for decades. However, the resulting geomorphic consequences, such as divide migration or valley infilling, are still poorly understood. Indeed, the prediction of the location of co-seismic landslides sources and deposits is a challenging issue and algorithms accounting for triggering of landslides by ground shaking are needed. Peak Ground Acceleration (PGA) has been shown to control at first order the spatial density of landslides. PGA can trigger landslides by two mechanisms: the direct effect of seismic acceleration on forces balance, and a transient decrease in hillslope strength parameters. To test the accuracy of modelling seismic waves effect by a cohesion drop, we use SLIPOS, an algorithm of bedrock landsliding based on a simple stability analysis (Culmann criteria) applied at local scale. The model is able to reproduce the landslide area-volume scaling relationship and the area-frequency distribution of natural landslides. Co-seismic landslide triggering is accounted for in SLIPOS by imposing a cohesion decrease. We calibrate the relationship between PGA and model cohesion via the landslides density. The spatial distribution of PGA is modeled using empirical relationships between earthquake source and PGA. We run the model on earthquake scenarios (Mw 6.5 to 8) applied to the Alpine fault of New-Zealand to predict the volume of landslides associated with large magnitude earthquakes. We demonstrate that simulating the effect of ground acceleration on landsliding using a cohesion drop leads to a realistic scaling between the volume of sediments and the earthquake magnitude.

International audience; The estimates of the worldwide sediment river discharge to the oceans suffer large uncertainties,mostly because sediment discharge measurements are scarce in space and time. Yet,knowing this sediment budget is critical in Earth and environmental Sciences, since it is underthe direct influence of surface processes (such as erosion, transport and sedimentation)and, in particular, controlled by climate, tectonics and human activities (dam, deforestation).Since 2002, the GRACE Satellite provides global gravity time series that have proven useful forquantifying mass transport, including continental water redistribution at the Earth surface (icesheets and glaciers melting, groundwater storage variations). Because, sediment accumulationin the oceanic sinks leads to a mass increase, we propose to use GRACE data to assess thesemass variations, and in turn the sediment discharge to the ocean. However, the order of magnitudeof mass variations due to sediments is only one tenth of the magnitude of continental water.As a result, unraveling the sedimentological contribution to GRACE signal is a challenge. Totackle it, we pair the analysis of regularized GRACE solutions at high spatial resolution correctedfrom all known contributions (hydrology, ocean, atmosphere) to a particle tracking modelthat predict the location of the sediment sinks for 13 rivers with the highest sediments loadsin the world. This particle tracking model combines global reanalysis of oceanic currents withavailable data on the grain size and the seasonality of the sediments load at the river mouthand is run over the same time range as GRACE data. We find encouraging correlation betweenGRACE spatial pattern of mass accumulation and the modeled pattern of sedimentation zonesfor some rivers, although some discrepancies between modeled and observed mass accumulationvalues are observed. The reasons for these misfits are a combination of inaccuracies bothin GRACE solutions and in the particle tracking model: 1) GRACE has a limited spatial resolutionand uncertainties, with residual contributions from off-target effects that are poorly estimated;2) The particle model is highly dependent on the oceanic currents’ velocity field, which alsosuffers from uncertainties and has limited spatial and temporal resolution as well; 3) We useliterature-based estimations of the seasonality of river sediment loads and associated grainsizes, both of which are notoriously uncertain data. We presently aim at refining our sedimentparticle tracking model in order to get better confidence in the spatial distribution of the sediments.This way, we expect to assess the annual riverine sediment discharge within the uncertaintyof GRACE, independently from uncertain in situ rivers’ sediment loads data.

International audience; Coseismic landsliding is an important contributor to the long-term erosion of mountain belts. But if the scalingbetween earthquakes magnitude and volume of sediments eroded is well known, the understanding of geomorphicconsequences as divide migration or valley infilling still poorly understood. Then, the prediction of the location oflandslides sources and deposits is a challenging issue. To progress in this topic, algorithms that resolves correctlythe interaction between landsliding and ground shaking are needed. Peak Ground Acceleration (PGA) have beenshown to control at first order the landslide density. But it can trigger landslides by two mechanisms: the directeffect of seismic acceleration on forces balance, and a transient decrease in hillslope strength parameters. Therelative importance of both effects on slope stability is not well understood. We use SLIPOS, an algorithm ofbedrock landsliding based on a simple stability analysis applied at local scale. The model is capable to reproducethe Area/Volume scaling and area distribution of natural landslides. We aim to include the effects of earthquakesin SLIPOS by simulating the PGA effect via a spatially variable cohesion decrease. We run the model (i) on theMw 7.6 Chi-Chi earthquake (1999) to quantitatively test the accuracy of the predictions and (ii) on earthquakesscenarios (Mw 6.5 to 8) on the New-Zealand Alpine fault to infer the volume of landslides associated with largeevents. For the Chi-Chi earthquake, we predict the observed total landslides area within a factor of 2. Moreover,we show with the New-Zealand fault case that the simulation of ground acceleration by cohesion decrease lead toa realistic scaling between the volume of sediments and the earthquake magnitude

International audience; In active mountain ranges, large earthquakes (Mw > 5-6) trigger numerous landslides that impact river dynamics.These landslides bring local and sudden sediment piles that will be eroded and transported along the river networkcausing downstream changes in river geometry, transport capacity and erosion efficiency. The progressiveremoval of landslide materials has implications for downstream hazards management and also for understandinglandscape dynamics at the timescale of the seismic cycle. The export time of landslide-derived sediments afterlarge-magnitude earthquakes has been studied from suspended load measurements but a full understanding of thetotal process, including the coupling between sediment transfer and channel geometry change, still remains anissue. Note that the transport of small sediment pulses has been studied in the context of river restoration, but themagnitude of sediment pulses generated by landslides may make the problem different. Here, we study the exportof large volumes (>106 m3) of sediments with the 2D hydro-morphodynamic model, Eros. This model uses a newhydrodynamic module that resolves a reduced form of the Saint-Venant equations with a particle method. It is coupledwith a sediment transport and lateral and vertical erosion model. Eros accounts for the complex retroactionsbetween sediment transport and fluvial geometry, with a stochastic description of the floods experienced by theriver. Moreover, it is able to reproduce several features deemed necessary to study the evacuation of large sedimentpulses, such as river regime modification (single-thread to multi-thread), river avulsion and aggradation, floods andbank erosion. Using a synthetic and simple topography we first present how granulometry, landslide volume andgeometry, channel slope and flood frequency influence 1) the dominance of pulse advection vs. diffusion during itsevacuation, 2) the pulse export time and 3) the remaining volume of sediment in the catchment. The model is thenapplied to a high resolution (5-10 m) digital elevation model of the Poerua catchment in New Zealand which hasbeen impacted by the effect of a large landslide during the last 15 years. We investigate several plausible AlpineFaults earthquake scenarios to study the propagation of the sediment along a complex river network.We characterizeand quantify the sediment pulse export time and mechanism for this river configuration and show its impact onthe alluvial plain evolution. Our findings have strong implications for the understanding of aggradation rates andthe temporal persistence of induced hazards in the alluvial plain as well as of sediment transfers in active mountainbelts.

International audience

International audience; The grain-scale morphology of river sediments and their size distribution are important factors controlling theefficiency of fluvial erosion and transport. In turn, constraining the spatial evolution of these two metrics offerdeep insights on the dynamics of river erosion and sediment transport from hillslopes to the sea. However, the sizedistribution of river sediments is generally assessed using statistically-biased field measurements and determiningthe grain-scale shape of river sediments remains a real challenge in geomorphology. Here we determine, with newmethodological approaches based on the segmentation and geomorphological fitting of 3D point cloud dataset,the size distribution and grain-scale shape of sediments located in river environments. Point cloud segmentation isperformed using either machine-learning algorithms or geometrical criterion, such as local plan fitting or curvatureanalysis. Once the grains are individualized into several sub-clouds, each grain-scale morphology is determinedusing a 3D geometrical fitting algorithm applied on the sub-cloud. If different geometrical models can be conceivedand tested, only ellipsoidal models were used in this study. A phase of results checking is then performed to removegrains showing a best-fitting model with a low level of confidence. The main benefits of this automatic method arethat it provides 1) an un-biased estimate of grain-size distribution on a large range of scales, from centimeter totens of meters; 2) access to a very large number of data, only limited by the number of grains in the point-clouddataset; 3) access to the 3D morphology of grains, in turn allowing to develop new metrics characterizing the sizeand shape of grains. The main limit of this method is that it is only able to detect grains with a characteristic sizegreater than the resolution of the point cloud. This new 3D granulometric method is then applied to river terracesboth in the Poerua catchment in New-Zealand and along the Laonong river in Taiwan, which point clouds wereobtained using both terrestrial lidar scanning and structure from motion photogrammetry

International audience; Climate cooling through the Late Cenozoic was important in the evolution of glaciated mountain ranges. Whilethe onset of accelerated Cenozoic exhumation is generally associated with the Quaternary at mid-latitudes, coincidentwith the local onset of glaciation, some high-latitude passive margins may have experienced earlier glaciationstarting at 30-38 Ma or even 45 Ma. To address this issue, we use a set of new AFT data from 16 sub-verticalprofiles sampled along the fjords of the central Eastern Greenland margin between 68° and 76°N, combined withnew apatite (U-Th-Sm)/He (AHe) data from selected profiles. To infer thermal histories and exhumation from theseprofiles, we use the software QTQt. The modeling results show a major phase of exhumation in the East Greenlandmargin between 68° and 76°N starting at 305 Ma. The spatial distribution of the exhumation shows that normalfaulting on East Greenland margin had no resolvable influence on exhumation related to the cooling phase. However,the timing is coincident with the dramatic worldwide fall of surface temperature at the Eocene-Oligocenetransition. We therefore suggest that a transition from an Eocene fluvial to an Oligocene glacial-dominated landscapetriggered a period of enhanced erosion. We infer from the thermal histories that around 2.71.9 km oferosion occurred close to the coast since the Eocene-Oligocene transition. This amount of erosion is consistentwith the incision of the fjords and with the effective removal of 2.31.5 km of basalt thickness, deduced by thethermal modeling of a heating phase at 555 Ma. This phase of erosion is most strongly evidenced near the coast,suggesting either that continental ice extent was limited to the coastal areas or that erosion was less efficient outsidethese areas, leading to no obvious signal in thermochronometric data further north. Overall, this study provides thefirst onshore evidence of the onset of continental ice in East Greenland margin at the Eocene-Oligocene transition(34 Ma), contemporary with the onset of Antarctica glaciation and erosion. Our interpretation is consistent withthat based on of the oldest ice-rafted debris found in the sedimentary records offshore the Eastern Greenland andimplies East Greenland exhibits the oldest onshore record of Cenozoic glacial erosion on Earth.

International audience; After the 2009 Morakot typhoon, which triggered numerous large landslides in Taiwan, Mouyen et al. (2013) showed for the first time the potential of time-lapse gravity survey to infer the mass of sediments transferring by landsliding or through rivers. By providing an integrated measurement of masses, gravimetry might thus be complementary to common methods used to assess the sediments discharge of rivers. But the masses of rocks displaced by Morakot were exceptionally large as a result of the record-breaking rainfalls brought by this typhoon and one might wonder to what extent time-lapse gravimetry could record such sediment transfers. In order to better assess the capabilities of this method, we set a time-lapse gravity network dedicated to the monitoring of such sediments transfers in Paolai village (south-central Taiwan). Paolai is located near the large Laonong river where temporary alluvial deposits of sediments exist and face steep mountain slopes likely to experience landslides. Both features are considered as potential source of mass transfers, and in turn of temporal gravity changes. The first base gravity measurements were done in November 2015, using absolute and relative gravimeters, and will be repeated every year, before and after the typhoon season. In the same time, we also use a terrestrial lidar to scan the geometry of both the river and the mountain slopes, hence providing a detailed topographical survey of the studied area. Adding Lidar measurements is an efficient strategy to solve for the non-uniqueness of gravity solutions. Meanwhile, we use the Eros morphodynamic model, that combine landsliding and flooding models, to investigate various scenarios of landsliding and subsequent sediment transport and compute the gravity changes on a virtual network of gravimeters. This gives us insights on the expected order of magnitudes for these surface sediment transfers, which are useful to unravel the induced gravity signal from others sources such as local hydrology or vertical ground displacements. The modeling results also enable us to set network location that are best defined to detect mass changes associated with sediment transfer. The main objective of this study is to test whether time-lapse gravity brings a new and reliable information compared to present methods monitoring sediments transfers and landscape evolution.

International audience

International audience

International audience