Séminaires
Laboratoire de Géologie
Date: June 30, 2026
Time: 11h
Location: salle Claude Froidevaux – E314
By: Daniel Douglas (Boston College)
Title: Linking Surface and Deep Earth Processes in Geodynamic Models: Coupling ASPECT and LandLab
Laboratoire de Géologie
Date: July 1, 2026
Time: 11h
Location: TBP (E409 ou E509)
By: David Marsan (Université Savoie Mont-Blanc)
Title: Friction models with cellular automata revisited: thermal activation, healing, and discolation models
Laboratoire de Géologie
Date: July 2, 2026
Time: 11h
Location: TBP (E409 ou E509)
By: Blandine Gardonio (CNRS/ISTerre Chambéry)
Title: Contrasted hydration state of the mantle-wedge corner revealed by its seismicity
Laboratoire de météorologie dynamique
Date: July 2, 2026
Time: 14h
Location: salle Claude Froidevaux – E314
By: Marc Bocquet (CEREA)
Title: Machine learning–driven advances in geophysical data assimilation
Abstract: Machine learning, and deep learning in particular, is becoming increasingly central to geophysical data assimilation, where it can enhance, complement, or even replace parts of the classical data assimilation cycle. I will illustrate recent advances in the field through two examples.
In the first, we investigate how deep learning can be used to discover new optimal analysis operators for sequential data assimilation applied to chaotic dynamics. I will show that the resulting “data assimilation networks” can address some of the most important challenges in classical data assimilation, such as properly representing the analysis uncertainty and accounting for departures of background statistics from Gaussianity.
In the second example, I will show how a generative AI-based surrogate model can be naturally incorporated into a classical yet advanced ensemble variational data assimilation scheme. The aim is to account for a broader class of model errors than those represented, for instance, in weak constraint 4D-Var.
Both cases will be illustrated through data assimilation experiments with low-order dynamics, potentially involving non-trivial model errors.
Laboratoire de Géologie
Date: July 3, 2026
Time: 11h
Location: salle Claude Froidevaux – E314
By: Marianne Métois (ENS Lyon)
Title: Tracking very slow deformation of continental Europe with GNSS and InSAR: challenges and perspectives
Laboratoire de Géologie
Date: July 7, 2026
Time: 11h
Location: salle Claude Froidevaux – E314
By: Annemiek Stegehuis (LGENS)
Title: From tidal pumping to tropical cyclones: controls on mangrove biogeochemistry across timescales
Abstract: Mangrove ecosystems are among the most carbon-rich coastal environments globally and play a key role in the exchange of carbon and nutrients between land and ocean. These exchanges are strongly influenced by hydrological forcing, ranging from daily tidal cycles to seasonal changes in freshwater inputs, and extreme events such as tropical cyclones.
In this seminar, I will present recent work on mangrove biogeochemistry across these different temporal scales. I will first discuss results from a 36-hour high-frequency sampling campaign in a tropical mangrove estuary in northeastern Brazil, where we investigated how tidal pumping and porewater exchange affect dissolved organic carbon (DOC), dissolved inorganic carbon (DIC), alkalinity, nutrients, and dissolved organic matter. I will then present ongoing seasonal studies in Brazil and The Gambia that aim to quantify how carbon cycling and outwelling vary throughout the year. Finally, I will examine the impact of tropical cyclones on mangrove carbon export, including recent evidence that extreme events can substantially reduce alkalinity, DIC and DOC outwelling and alter coastal carbon dynamics.
Together, these studies provide new insights into the mechanisms controlling carbon and nutrient exchange between mangrove ecosystems and the coastal ocean.
Laboratoire de météorologie dynamique
Date: July 9, 2026
Time: 11h
Location: salle Claude Froidevaux – E314
By: Martina Kraemer (Université de Mainz)
Title: Environmental Influences on Ice Crystal Sizes in Cirrus Clouds
Abstract: Cloud particle size distributions (PSDs) are crucial in determining the clouds physical and optical properties and hence their radiative feedback to the climate. Here we present unprecedented occurrence patterns of ice crystals in cirrus clouds derived from 270 h of cloud measurements (≈ 975 000 PSDs).
In particular, cirrus PSDs for cold to warm cirrus temperatures and microphysically thin to thick cirrus clouds are provided in a novel presentation as heat maps. The observations are accompanied by simulations of ice crystal growth in in-situ-origin cirrus.
Larger crystals most likely originate from clouds that formed at lower altitudes and rose into the cold cirrus layers, where the cloud particles glaciate. The combined evaluation of observations and simulations allows the attribution of processes shaping the PSDs.
Important results are that, as temperature and cirrus thickness decrease, the prevailing ice particle population evolves from larger and more abundant crystals characteristic of liquid-origin cirrus to few, small crystals typical of in situ-origin cirrus. This shift in cirrus origin is associated with substantially different climate feedbacks.
Laboratoire de Géologie
Date: July 9, 2026
Time: 14h
Location: salle Claude Froidevaux – E314
By: Pathikrit Bhattacharya (NISER)
Title: Is the collective slip behavior of faults universal? Comparing plate boundaries and plate interiors
Abstract: Over the last two decades, it has become clear that plate boundary faults slip to release stored strain energy not only through earthquakes but instead a continuum across time and slip rate scales including aseismic slow slip, low and very low frequency earthquakes. One explanation for how these diverse modes of slip events occur appeals to variations in frictional properties mainly along dip of plate boundary fault zones related to differences in mineralogy, temperature, pressure and deformation regimes across the vast depth range (5 to 60 km or more) of these slip events. For physical conditions relevant to subduction zones, laboratory experiments largely suggest that such an explanation is plausible.
In this talk, I will show you results from a shallow crustal ‘earthquake swarm’ within the Deccan Volcanic Province in Western India where aseismic slip spanning months, low-frequency earthquakes and regular earthquakes all happen within a depth range of 1-6 km on steeply dipping normal faults. Remarkably, these are the first observations showing that the entire spectrum of fault slip, normally linked to plate boundary faults, also occurs within the plate interior. Given their severely depth-constrained occurrence, these observations seem to suggest that variations in frictional properties along a fault zone might not be a ‘universal’ requirement for such diverse modes of fault slip to emerge. Instead, I will show that the temporal evolution of both aseismic slip and seismicity within this swarm provides direct evidence for fault interactions between the principal faults and their damage zones. These findings seem to support the recent theoretical prediction that mechanical interactions between fault networks and their associated damage zones are sufficient to produce the full spectrum of fault slip behaviors observed at plate margins and might represent a ‘universal’ collective slip behavior of fault zones.
Laboratoire de Géologie
Date: September 15, 2026
Time: 11h
Location: salle Claude Froidevaux – E314
By: Marouchka Froment (IPGP)
Title: TBA
Laboratoire de Géologie
Date: September 29, 2026
Time: 11h
Location: salle Claude Froidevaux – E314
By: Eric Larose (ISTerre)
Title: TBA
Past seminars
Soutenance d’HDR
Date: June 25, 2026
Time: 14h
Location: salle Claude Froidevaux – E314 et en visio ici
By: Hugo Bellenger (LMD)
Title: Échanges d’énergie et de matière entre l’océan et l’atmosphère : du millimètre à l’échelle globale
Abstract: Il est bien connu que l’atmosphère et l’océan échangent en permanence de l’énergie et de la matière et que ces échanges sont déterminants pour les phénomènes météorologiques et pour l’évolution du climat. Ces échanges induisent directement des perturbations de la dynamique de l’océan et de l’atmosphère, mais ils interagissent également avec l’état de la fine couche océanique de l’interface. Ainsi, un coup de vent forçant une onde de Kelvin équatoriale va déclencher un événement El Niño et engendrer un mélange avec les eaux plus profondes, mais il va aussi accroître les vagues et la turbulence qui modifient les caractéristiques de l’interface. L’état de l’interface va à son tour influer sur les échanges d’eau et d’énergie qui nourrissent la convection profonde et les phénomènes tropicaux tels que moussons, oscillation de Madden-Julian et cyclones. Il va également modifier les cycles biogéochimiques et en particulier la capacité de l’océan à absorber le CO2.
Du fait de la difficulté de son observation directe, la physique de l’interface reste mal connue. Elle est cependant intéressante car elle a des signatures significatives sur les paramètres qu’il est possible d’observer à l’échelle globale depuis l’espace comme la température, la salinité ou la rugosité de la mer. C’est pour interpréter correctement ces signatures qu’il est nécessaire d’étudier les processus de fines échelles (du millimètre à quelques mètres) impliqués dans les échanges air-mer. Je présenterai les processus à l’œuvre au voisinage direct de l’interface comme la diffusion moléculaire, la turbulence, l’atténuation des ondes capillaires par les pellicules de molécules tensio-actives d’origine naturelle. Je montrerai comment ils sont influencés localement par le forçage atmosphérique, en particulier par les flux radiatifs et la pluie, et quelle est leur influence sur la température, la salinité, la rugosité et sur les échanges de gaz comme le CO2. Enfin, j’évaluerai leur importance sur le climat global et son observation.
Jury: Jacqueline, BOUTIN, LOCEAN, Présidente
Rémy, ROCA, LEGOS, Rapporteur
Karine, SELLEGRI, LAMP, Rapportrice
Jean-Luc, REDELSPERGER, LOPS, Rapporteur
Laurent, BOPP, DR CNRS, LMD, Examinateur, Correspondant ENS
Département de Géosciences
Date: June 19, 2026 – FRIDAY
Time: 11h
Location: salle Claude Froidevaux – E314
By: Raphaël Grandin (IPGP)
Title: Near-real-time remote sensing of volcanic activity from space
Laboratoire de météorologie dynamique
Date: June 18, 2026
Time: 11h
Location: salle Claude Froidevaux – E314
By: Takahito Kataoka (JAMSTEC)
Title: Seasonal to decadal climate prediction with MIROC6
Abstract: The study presents results of seasonal‐to‐decadal climate predictions using a coupled climate model called the MIROC6, which contributed to the CMIP6. MIROC6 is initialized every year for 1960–2018 by assimilating observed ocean temperature and salinity anomalies and full-fields of sea ice concentration and by prescribing atmospheric initial states from reanalysis data. The impacts of updating the system on prediction skill are then evaluated by comparing hindcast experiments between the MIROC6 prediction system and the previous system based on MIROC version 5 (MIROC5). The skill of seasonal prediction is overall improved in association with representation and initialization of El Niño/Southern Oscillation (ENSO), the Quasi‐Biennial Oscillation (QBO), and the Northern Hemisphere sea ice concentration in MIROC6. In particular, the QBO is skillfully predicted up to 3 years ahead with a maximum anomaly correlation exceeding r = 0.8. The prediction skill for the North Atlantic Oscillation in winter is also enhanced, but the prediction still suffers from the model’s inherent errors. On decadal timescales, MIROC6 has a larger fraction of areas of the globe with better surface temperature skill at all lead times than MIROC5, and it has predictive skill in the annual‐mean sea surface temperature (SST) in the North Atlantic and the Pacific. In particular, MIROC6 hindcasts at 2–5 years lead time are able to capture the spatial structure of SST changes in the North Pacific and the eastern tropical Pacific associated with the 1970s regime shift better than MIROC5 hindcasts. The impacts of a large ensemble are also discussed.
In addition, I will briefly highlight our efforts on earth system predictions and mechanism studies on tropical climate variability.
Département de Géosciences
Date: June 17, 2026
Time: 11h
Location: salle Claude Froidevaux – E314
By: Eulalie Boucher (ECMWF)
Title: L’utilisation de l’intelligence artificielle pour la prevision numérique du temps au CEPMMT
Abstract: Cette présentation donnera un aperçu général de l’utilisation du Machine Learning au sein du CEPMMT. Nous présenterons le système opérationnel AIFS (AI Forecasting System), premier système de prévision basé sur l’intelligence artificielle déployé de manière opérationnelle au CEPMMT, ainsi que ses performances et ses perspectives d’évolution. La présentation abordera également les travaux en cours autour d’un futur système fondé uniquement sur les observations (DOP – Direct Observation Prediction), ouvrant la voie à une nouvelle génération de modèles de prévision météorologique.
Laboratoire de Géologie
Date: June 16, 2026
Time: 11h
Location: salle Claude Froidevaux – E314
By: Soumaya Latour (IRAP-OMB)
Title: Direct estimation of rupture properties from a single CCTV camera, Mw7.7 Mandalay 2025 earthquake
Abstract: We present the analysis of the first known video of co-seismic slip on a natural fault. It was captured during the 2025 Mw 7.7 Mandalay earthquake (Myanmar) by a CCTV camera located a few meters away from the fault. By direct image analysis of the footage, we measure the slip and slip-rate functions from a natural coseismic rupture. The results show that the rupture propagated as a slip-pulse, with a local slip duration of 1.4 s and a maximum slip rate of 3.5 m/s ± 20%. We then fit two steady-state slip-pulse models to the measured slip-rate function, allowing us to estimate the mechanical properties of the fault: the slip-stress curve, the slip-weakening distance, the breakdown work and the energy release rate. This study shows the value of direct on-fault slip measurements for constraining those mechanical parameters, that are key inputs in dynamic rupture models.
Soutenance d’HDR
Date: June 15, 2026
Time: 15h
Location: salle Claude Froidevaux – E314 et en visio ici
By: Nicolas Rochetin (LMD)
Title: Thermiques, brises, courants de densité et nuages : la convection déconstruite.
Du développement de paramétrisations pour LMDZ à la campagne aéroportée MAESTRO
Abstract: Depuis le monde des paramétrisations de la convection pour les modèles à grosses grilles jusqu’aux observations in-situ, en passant par les modèles explicites à aire limitée, mon parcours scientifique m’a convaincu de la pertinence et de la nécessité de conduire en parallèle le travail de construction (développement de modèle à grosses grilles) et de déconstruction (analyse de sorties de simulations à haute résolution) de la convection, ici appréhendée comme une somme d’objets (ou processus) couplés entre eux et à la circulation de grande échelle. A travers l’exploration des thermiques de couche-limite, des circulations à méso-échelle et des courants de densité et de leurs couplages avec la convection profonde (les orages), mes travaux sur le cycle de vie et l’organisation spatiale de la convection démontrent qu’en plus d’éclairer notre interprétation des données issues de simulations numériques et d’observations de dernière génération, les paramétrisations de la convection permettent aussi, par l’addition de lois empiriques et l’introduction de paramètres nouveaux, d’inspirer des stratégies de mesure pour les évaluer et favorisent l’émergence de nouvelles théories permettant de donner un sens physique à tout l’ensemble.
Aussi bien au niveau national qu’international, la convection atmosphérique est une discipline en pleine effervescence dont les dernières avancées laissent augurer la percée prochaine de théories nouvelles qui bousculeront un certain nombre de paradigmes hérités des premières heures de la modélisation de la convection et du climat. En la matière, même si la haute résolution et l’IA ouvrent indéniablement des perspectives prometteuses, ces approches ne nous permettent pas (encore) de nous affranchir de la nécessité de proposer des représentations conceptuelles de la convection pour avancer dans sa compréhension et sa modélisation. En ce sens, les paramétrisations de la convection ont donc encore de beaux jours devant elles, et elles rempliront peut-être même un rôle déterminant dans la gestation et l’avènement de ces nouvelles théories.
Jury:
Rapporteurs : Steve Sherwood (UNSW, Sydney), Sylvie Malardel (Météo France), Gerhard Krinner (IGE)
Examinateurs : Fleur Couvreux (CNRM/Météo France), Fabienne Lohou (CRA, Univ. Paul Sabatier), Freddy Bouchet (LMD)
Mentor : Sandrine Bony (LMD)
Département de Géosciences
Date: June 12, 2026
Time: TBA
Location: salle Claude Froidevaux – E314
By: Journée du département
Laboratoire de Géologie
Date: June 09, 2026
Time: 11h
Location: salle Claude Froidevaux – E314
By: Carolina Giorgetti (LGENS)
Title: The role of the tectonic stress field and fault damage zone in earthquakes: bridging laboratory and field perspectives
Bureau des longitudes
Date: June 3, 2026
Time: 14h30
Location: salle Dussane – 45 rue d’Ulm
By: Julien Aubert (CNRS – Institut de physique du globe de Paris)
Title: Comprendre le signal géomagnétique, de l’année au milliard d’années
Département de Géosciences
Date: May 26, 2026
Time: 11h
Location: salle Claude Froidevaux – E314
By: Réunion plénière du LRC Yves Rocard
Programme: TBA
Laboratoire de Géologie
Date: May 19, 2026
Time: 11h
Location: salle Claude Froidevaux – E314
By: Olivier Galland (UiO)
Title: Beyond Elasticity – How inelastic properties of crustal rocks control the propagation of dykes and sills in volcanic plumbing systems
Laboratoire de météorologie dynamique
Date: May 18, 2026
Time: 14h
Location: salle Claude Froidevaux – E314
By: Joseph Mouallem (GFDL)
Title: Advances in Grid Algorithms for Weather and Climate Models: From Local Refinement to Global Coupling
Abstract: Simulating multiscale geophysical flows requires numerical frameworks that balance accuracy, efficiency, and scalability across a wide range of spatial scales. This presentation highlights recent advances in grid design and model coupling within next-generation Earth system models at GFDL/NOAA. I will present the Duo-Grid approach in NOAA’s dynamical core FV3 for mitigating grid imprinting on the cubed-sphere, improving numerical accuracy by ensuring physically consistent cross-edge data representation. I will then discuss two-way grid nesting strategies that enable targeted high-resolution refinement while maintaining consistency with the global solution. Finally, I will introduce recent developments in high resolution coupled atmosphere-ocean-land modeling, that enable consistent two-way interactions across model components. Together, these advances provide a unified framework for improving the fidelity and scalability of high-resolution simulations, with applications to extreme weather forecasting and climate prediction.
Laboratoire de Géologie
Date: May 11, 2026 (MONDAY)
Time: 11h
Location: salle Claude Froidevaux – E314
By: Gina McGill (ISTO)
Title: Micro-porosity in quartz mylonites
Bureau des longitudes
Date: May 6, 2026
Time: 14h30
Location: salle Dussane – 45 rue d’Ulm et en visio ici
By: Christian Bizouard (Observatoire de Paris – Laboratoire Temps Espace)
Title: Etudier la Terre en auscultant sa rotation
Abstract: La rotation de la Terre n’est pas tout à fait uniforme. A l’échelle d’une année, la vitesse de rotation varie d’un dix-millionième, l’axe de rotation se déplace d’une minute de degré par rapport aux étoiles et d’une dizaine de mètres sur la croûte terrestre.
Ces variations affectent l’orientation du référentiel terrestre par rapport au référentiel céleste, et doivent être considérées pour le pointé des télescopes ou le suivi des satellites ou des sondes spatiales. Ce que l’on sait moins, c’est que ces fluctuations permettent aussi de déterminer les propriétés notre planète et les processus qui l’affectent.
C’est cette aventure intellectuelle, initiée au 18e siècle, et révolutionnée par les techniques astro-géodésiques modernes, dont vous voulons donner un aperçu.
En particulier nous verrons comment l’observation de la précession-nutation permet de sonder l’intérieur de la Terre, comment les effets réguliers des marées renseigne la rhéologie terrestre, dans quelle mesure les irrégularités dans le mouvement du pôle et la vitesse de rotation valident notre connaissance de la circulation hydro-atmosphérique, des tendances climatiques, et d’autres processus géophysiques.
Certaines variations n’en demeurent pas moins incomprises ou imprévisibles, faisant de la rotation terrestre un domaine à la fois ouvert et excitant
Laboratoire de Géologie
Date: April 29, 2026
Time: 11h
Location: salle Claude Froidevaux – E314
By: Charles G. Sammis (University of British Columbia)
Title: Mechanics and Scaling of Episodic Tremor and Slip in Northern Cascadia
Département de Géosciences
Date: April 28, 2026
Time: 11h
Location: salle Claude Froidevaux – E314
By: Groupe FLORA led by Carole Dalin (LGENS)
Effect of global diet change on environmental sustainability
par Marcellin Guilbert
By coupling our 2020 assessment of cropland environmental sustainability with an input–output model (essentially a recipe book of global production), we assess how shifts in global diets affect sustainability.
Characterisation of global cropland bright spots
par Jasmine Gamblin
We mapped the distribution of « bright spots » in global agriculture—regions where significant crop production does not exceed local environmental sustainability thresholds. Studying the shared characteristics of these regions using random forest classification highlights the importance of natural habitat around cropland. We thus explore the possibility of extending sustainable regions by rewilding some fraction of cropland, and quantify the associated production loss.
Consumption-based GHG footprint of global food systems (2000–2020)
par Belén Benitez
Food consumption drives environmental pressures by shaping global agricultural production systems and international trade patterns. Developing harmonized frameworks that consistently integrate multiple agricultural GHG emission sources and link them to food consumption through trade is therefore essential for fully assessing the sustainability of the agri-food system. Here we quantify the carbon footprint of food consumption by combining spatially-explicit land use change-related GHG emissions. Results are reported at the global scale across four reference years between 2000 and 2020. By reallocating production-based emissions to final consumers through a consumption-based framework, we link global food demand to the geographic origin of agricultural GHG emissions, thereby enabling an analysis of spatial patterns and temporal trends of the carbon footprint of food demand worldwide
Département de Géosciences et IBENS
Date: April 27, 2026 (MONDAY)
Time: 11h
Location: salle Claude Froidevaux – E314
By: Scott Saleska (Univ of Arizona)
Title: From Landscape Change to Planetary Terraformation: Advancing a convergent “Landscape Terraformation science” of how life transforms planets with a multi-scale collaboratory digital twinning of Biosphere 2
Abstract: Humanity faces a dual planetary challenge: to restore Earth’s rapidly degrading landscapes, which threaten biodiversity, water, and food security, and to develop the foundational science for terraforming extraterrestrial planets, which may one day extend the resilience of life beyond Earth. These challenges are deeply connected, grounded in the shared need to understand how life transforms barren environments into habitable ecosystems—what we call here the challenge of “landscape terraformation”. Advancing Landscape Terraformation science as a new solution to this dual challenge requires predictive, adaptive systems that navigate complexities of ecosystem emergence across radically different conditions. With the advent of AI models, real-time sensing, and digital twin technologies, we now have unprecedented ability to couple experimentation with intelligent computation. Thus, we here review initial examples of a multi-scale dynamic “collaboratory” that learns by integrating real-world experimentation and digital twinning of UA’s unique Landscape Evolution Observatory (LEO) at Biosphere 2, in order to predict, trigger, and steer terraformation processes.
Scott Saleska is an ecologist who investigates how life reshapes Earth’s surface and climate—from tropical forests to engineered landscapes. His work combines experiments, remote sensing, and digital twinning technologies to understand and predict planetary-scale change. He is a professor of ecology and evolutionary biology, Biosphere 2 Director of Science, and Co-Director of a graduate training program in Ecosystem Genomics, all at the University of Arizona. He is elected fellow of the American Geophysical Union and of the Ecological Society of America; he was a visiting academic at ENS (2019), and will be a U.S.-Brazil Fulbright Scholar in 2026-2027.
Laboratoire de Géologie
Date: April 21, 2026
Time: 11h
Location: salle Claude Froidevaux – E314
By: Mark Behn (Boston College)
Title: Rheologic constraints on the thermal structure and seismogenic behavior of oceanic transform faults
Laboratoire de météorologie dynamique
Date: April 8, 2026
Time: 10h
Location: salle Claude Froidevaux – E314
By: Alex Wikner (UChicago)
Title: AI Climate Emulators for Predicting Extreme Weather Statistics
Abstract: Estimating the risk of extreme weather due to climate change, particularly at regional scales, is critically important yet remains one of the most challenging problems to address with traditional physics-based global climate models (GCMs). The computational cost of these models leads to high uncertainty in estimates of the rarest — yet most impactful — extreme weather return periods. AI emulators trained on historical reanalysis data have been shown to reproduce global atmospheric dynamics at greatly reduced computational cost and, in some cases, with decades-long stability. Such emulators can also be combined with coarse-resolution physics-based models to simulate extreme weather events that are absent from the physics-based model alone. I will discuss this in the context of an early reservoir computer-based emulator that successfully simulates sudden stratospheric warming events not captured by its component physics-based model.
However, because emulators are trained and validated on the relatively short historical record, it is difficult to assess whether their predicted extreme event statistics are accurate. Using state-of-the-art emulator architectures, we performed a first-of-its-kind assessment of this capability using 92,000 years of stationary climate data. We find that emulators can generate weather events more extreme than those in the training set that are dynamically similar to those in the long simulation data. However, the accuracy with which emulators reproduce the correct return periods of the most extreme events varies by region, variable, and architecture. These biases are difficult to correct with conventional training methodologies, motivating the development of new sampling methods. One such method, AI-boosted rare event sampling, uses emulator forecasts as a score function in an importance sampling algorithm to more efficiently sample extremes from the physics-based model and produce unbiased return period statistics. I will conclude by discussing future prospects for resampling approaches to improve emulated extreme event statistics.
Laboratoire de Géologie
Date: April 7, 2026
Time: 11h
Location: salle Claude Froidevaux – E314
By: Nico Bigaroni (Manchester Univ)
Title: Seismic–Aseismic Slip Partitioning on a Frictionally Heterogeneous Fault: An Experimental Approach
Abstract: Natural faults are frictionally heterogeneous, commonly comprising velocity‑weakening (VW) patches embedded within velocity‑strengthening (VS) regions, a configuration thought to control earthquake nucleation and rupture dynamics. While this framework can be inferred from exhumed faults and seismology models, experimental tests using realistic patch geometries and rock samples remain scarce.
We present a new experimental approach to investigate heterogeneous fault slip using decimetre-scale rock samples sheared in the newly developed BeeAx servo-controlled biaxial apparatus. Pennant Sandstone blocks (15 × 17 cm) were deformed at ~1 µm/s under 2–8 MPa normal stress in three configurations: homogeneous sandstone (VW), homogeneous graphite-coated sandstone (VS), and heterogeneous faults comprising four circular sandstone patches embedded in a graphite background (50% of the sliding surface). Sixteen calibrated acoustic emission (AE) sensors enabled event location and source parameter analysis.
Homogeneous graphite samples exhibited stable sliding with low friction (µ ≈ 0.15), whereas homogeneous sandstone showed unstable stick-slip with higher friction (µ ≈ 0.5). Heterogeneous samples displayed hybrid behaviour, with low friction (µ ≈ 0.20) comparable to graphite but persistent stick-slip events. AE hypocentres localized around the perimeters of sandstone patches, indicating stress transfer from creeping graphite to locked VW domains. Compared to homogeneous sandstone, heterogeneous faults showed larger stress drops, stronger seismic localization, and systematically lower Gutenberg–Richter b-values. Temporal b-value evolution was constant for graphite, cyclic for sandstone, and intermediate but suppressed in heterogeneous samples, consistent with enhanced patch interaction.
These results demonstrate that weak VS regions can actively load and amplify seismic slip-on stronger VW patches, providing experimental support for patch-controlled seismicity on natural faults and in induced seismic settings. More broadly, this framework enables controlled investigation of rupture nucleation and stress interaction in frictionally heterogeneous fault systems.
Bureau des longitudes
Date: April 1st, 2026
Time: 14h30
Location: salle Dussane – 45 rue d’Ulm et en visio ici
By: Yves Marrocchi (Centre de Recherches Pétrographiques et Géochimiques / CRPG-CNRS, Nancy)
Title: Aux origines du Système solaire : ce que nous apprennent les retours des astéroïdes Ryugu et Bennu
Abstract: Comprendre la formation et l’évolution du système solaire constitue l’un des grands objectifs des sciences planétaires. Dans cette présentation, je décrirai les différents objets et méthodes qui permettent d’explorer les premières étapes de son histoire, depuis l’étude des météorites jusqu’aux observations astronomiques et aux expériences de laboratoire. Je dresserai ensuite un état des lieux de notre compréhension actuelle de la formation et de l’évolution du système solaire, en insistant sur les processus qui ont gouverné l’assemblage des premiers solides et la formation des planètes.
Je présenterai également les récentes missions spatiales de retour d’échantillons d’astéroïdes, notamment celles qui ont rapporté sur Terre des fragments des astéroïdes Ryugu et Bennu. Ces échantillons offrent une opportunité unique d’étudier des matériaux primitifs du système solaire dans des conditions analytiques exceptionnelles. Je présenterai les principales découvertes issues de leur analyse, en particulier celles concernant la dynamique des solides ainsi que la distribution de l’eau et de la matière organique dans le disque protoplanétaire.
Département de Géosciences
Date: March 31, 2026
Time: 11h
Location: salle Claude Froidevaux – E314
By: Kristel Chanard (IGN/IPGP)
Title: The deforming Earth as a sensor of the changing water cycle
Abstract: Climate change and human activities are profoundly reshaping the Earth’s water cycle, driving accelerated ice sheet and glacier melt, increasing precipitation variability, intensifying evaporation, accelerating groundwater depletion, and contributing to rapid sea level change. Understanding these transformations requires integrating complementary observations of water mass redistribution to improve predictions and support evidence-based decision-making for water and coastal risk management.
In this context, modern space geodesy, which measures changes in the Earth’s gravity field and surface deformation, provides a unique way to monitor the redistribution of water across continental hydrology, the cryosphere, and the oceans over spatial scales from local to global and timescales from seasonal to multi-decadal. However, interpreting these observations in terms of water, ice, and ocean redistribution remains challenging due to the integrated nature of geodetic signals.
We will explore the potential of combining multiple geodetic observations, including spatio-temporal variations in the Earth’s gravity field and surface deformation measured by GNSS and InSAR, with in-situ measurements and physics-based models of the solid Earth-water interactions to disentangle solid Earth processes from hydrological, cryospheric, and oceanic signals. This approach not only helps refine estimates of water and ice mass redistribution from space geodesy, but also provides new insights into the mechanical properties of the solid Earth, with implications for a better understanding of other geophysical processes.
Laboratoire de météorologie dynamique
Date: March 26, 2026
Time: 11h
Location: salle Claude Froidevaux – E314
By: Olivier Pauluis (NYU)
Title: Atmospheric overturning, from the convective to the planetary scales
Abstract: The atmosphere absorbs solar energy mostly at the Earth’s surface and in low latitudes, and loses energy by emitting infrared radiation from the troposphere. This differential heating drives an atmospheric circulation that redistributes the energy around the globe. This circulation is an overturning flow characterized by the ascent of warm, moist air balanced by the subsidence of colder, drier air. It also involves a broad range of scales, from the global planetary Hadley cell to individual convective clouds.
In this talk, I will show how atmospheric overturining can be quantified using isentropic analysis. This method relies on separating air parcels through their equivalent potential temperature. The overturning circulation then appears as an upward mass flux of air at a high equivalent potential temperature, balanced by a descending mass flux of air at a low equivalent potential temperature. This method has been successfully applied to identify the overturning in high-resolution simulations of convection, tropical cyclones, and the Indian Summer Monsoon. When applied to a global dataset, one can also determine the contributions of different motion scales to the overturning.
A significant challenge to our understanding of atmospheric overturning is that, to date, no global dataset has fully resolved all the scales involved. However, the emergence of Global Cloud Resolving Models (GCRMs) such as XSHIELD and ICON makes it now possible to investigate the contribution of convective motions to this overturning directly. In particular, because they resolve convective motion, these models produce an overturning circulation that is about twice as strong as that of the previous generation of global models.
Laboratoire de Géologie
Date: March 25, 2026 (mercredi)
Time: 11h
Location: salle Claude Froidevaux – E314
By: Mohsen Talebkeikhah (EPFL)
Title: Closure Analysis of Hydraulic Fractures with a Novel Aperture Measurement and Micro-Scale Analysis of Residual Fracture
Abstract: This study investigates the propagation and closure behavior of hydraulic fractures in a cubic block of porous Molasse sandstone under multiple cycles of fluid injection and shut-in. Using an electromagnetic induction-based Eddy Current (EC) probe, we measured real-time fracture opening. The experimental setup included variations in fluid viscosity and confining stresses, capturing fracture responses across several propagation and closure cycles. Our findings reveal that fracture behavior primarily aligns with the M and M’ vertex regimes, with fluid storage and leak-off significantly influencing the fracture evolution. Following this, we performed closure analysis and observed the semi-analytical Sunset solution at peak closure time and pressure. Notably, 3D reconstruction of the residual fracture surface, using CT scans of core samples in the fracture region, reveals that fractures do not close entirely, leaving a residual opening. Micro-scale roughness analysis captured the contact point distribution and yielded a hurst exponent value of 0.4, consistent with literature. This confirmed that fractures propagate intergranularly, surrounding the particles, as observed in the CT images.
Soutenance de THÈSE
Date: March 23, 2026
Time: 11h
Location: amphithéâtre du Pôle Mécanique, bâtiment 104, avenue Becquerel, 91120 Palaiseau et en visio ici
By: Matthieu Lusseyran (LMS / LGENS)
Title: Mécanismes de microdéformation des roches argileuses mis en évidence par la vitesse des ondes P et la corrélation d’images numériques
Abstract: La compréhension des processus d’endommagement qui gouvernent le comportement mécanique des roches argileuses constitue un enjeu majeur en géotechnique et en géo-sismique. Ces roches jouent un rôle déterminant dans la stabilité des systèmes naturels, où elles contrôlent la localisation de la déformation et la rupture de zones de faille, mais également dans les environnements géologiques aménagés. En raison de leur très faible perméabilité, de leur capacité de cicatrisation des fractures et de leur forte affinité pour les radionucléides, plusieurs pays européens envisagent les formations argileuses comme roches hôtes pour le stockage géologique profond des déchets radioactifs. Dans ce contexte, l’intégrité de la roche hôte est cruciale, car tout endommagement peut accroître le transport advectif et compromettre la sûreté du stockage. La détection précoce des dommages et l’identification des mécanismes qui gouvernent leur évolution constituent donc un enjeu central.
Les méthodes d’auscultation non destructives, et en particulier la propagation des ondes ultrasonores, sont efficaces pour suivre l’évolution interne des roches. En effet, les variations de vitesses des ondes P et S sont sensibles aux changements microstructuraux, tels que la fermeture des pores, l’initiation ou la croissance de fissures et les réorganisations plastiques. Elles sont donc utilisées pour détecter la localisation des contraintes, suivre la progression de l’endommagement et anticiper la rupture, que ce soit dans des zones de faille ou dans des ouvrages souterrains. Toutefois, l’interprétation de ces signatures reste délicate, car les mécanismes ductiles et fragiles peuvent produire des réponses acoustiques similaires.
Des études en laboratoire ont recours à l’imagerie haute résolution et à la corrélation d’images (DIC) pour analyser la microdéformation des roches argileuses. Ces travaux ont mis en évidence une grande variété de mécanismes telles que la réorientation des feuillets, glissement intergranulaire, compaction localisée ou coalescence de microfissures, mais leur influence sur les vitesses ultrasonores et l’anisotropie élastique reste discutée. Peu d’expériences permettent en effet de suivre en parallèle l’évolution microstructurale et la réponse élastique dynamique.
Cette thèse s’attache à combler cette lacune en étudiant le couplage entre l’évolution des vitesses d’ondes P et les micromécanismes de déformation identifiés par DIC dans l’argilite de Tournemire. Un dispositif expérimental original a été développé au Laboratoire de Mécanique des Solides (LMS). De petits échantillons sont soumis à une compression uniaxiale sous humidité contrôlée, tandis que deux mesures synchrones sont réalisées : les temps de propagation ultrasonore axiaux et latéraux, et les champs de déplacement obtenus par DIC à partir d’images optiques et de microscopie électronique environnementale (ESEM). Cette approche multi-échelle permet d’identifier les zones de localisation et de coupler les réarrangements microstructuraux avec l’évolution de la vitesse des ondes P.
En couplant acoustique et imagerie au cours d’un même essai, cette méthodologie établit un lien direct entre la nucléation du dommage, sa propagation et l’évolution de l’anisotropie élastique. Les résultats apportent de nouveaux éléments sur les micro mécanismes de déformation dans les roches argileuses, la sensibilité des vitesses ultrasonores aux transformations microstructurales et le potentiel de ces observations couplées pour améliorer le suivi géophysique des formations argileuses naturelles ou aménagées.
Laboratoire de météorologie dynamique
Date: March 20, 2026
Time: 14h
Location: salle Claude Froidevaux – E314
By: Robin Noyelle (ETH Zürich)
Title: Megadroughts and megapluvials in CESM2 arise from atmospheric stochasticity rather than ocean-atmosphere coupling
Abstract: Megadroughts and megapluvials are multi-year dry and wet events of exceptional intensity and duration. There is strong paleoclimate evidence for the existence of such events in various regions of the globe. However, the mechanisms how these anomalies can be sustained for long periods of time have not been elucidated and reproduced in climate models. Here we address this question using large-ensemble simulations with the CESM2 fully-coupled climate model. We argue that, outside ENSO-influenced regions, meteorological megadroughts and megapluvials in the simulations are mainly caused by natural variability in the atmosphere with limited influence from the oceans. We first show that interannual correlations in accumulated precipitation are weak: multi-year extreme precipitation events arise as a succession of independent yearly events. Secondly, anomalous SST patterns do not explain the intensity of dry and wet years. Finally, we propose a simple, purely stochastic model of atmospheric precipitation that can reproduce most of the key characteristics of simulated extreme dry and wet years. These results imply that, outside ENSO-influenced regions, megadroughts and megapluvials in the climate model are caused by the succession of independent dry and wet years, which are themselves stochastically expected large deviations arising from the natural variability of the atmosphere, and not from the interaction with the slower ocean. However, the intensity and frequency of recorded megadroughts and megapluvials are not compatible with these model results. This strongly suggests that key physical mechanisms are missing in the model to reproduce these peculiar events and advocates for caution in estimating the probability of multi-year dry and wet events from climate model simulations.
Laboratoire de Géologie
Date: March 17, 2026
Time: 11h
Location: salle Claude Froidevaux – E314
By: Théo Briolet (LGENS)
Title: Controls of double-porosity interconnectivity on dissolution of carbonate rocks
Abstract: Carbonate rocks are characterized by an important reactivity to fluid circulation, leading to dissolution phenomena associated to significant changes in their hydromechanical (petrophysical and mechanical) properties. These processes play a key role in the genesis of karstic environments, which constitute major reserves of drinking water, and must be integrated to ensure large-scale deployment of geological storage of CO2 in carbonate saline aquifers.
Dissolution of carbonate rocks is controlled by different factors, related to the fluid (chemical composition, pH, flow velocity) or the rock (mineral content, porosity, permeability, microstructure), and results in different dissolution patterns. Even though microstructural controls on dissolution processes have been previously evidenced, the specific contribution of the double porosity (i.e., the intergranular macroporosity and intragranular microporosity, commonly found in carbonate rocks) remains unclear, and is so far poorly considered in experimental and numerical studies. To investigate the specific impact of the double-porosity interconnectivity on the dissolution of carbonate rocks, we perform controlled dissolution experiments under identical experimental conditions on two nearly pure calcite limestones (Euville and Lavoux) characterized by different grain and pore structures. Two flow rates are applied to investigate the effect of various hydrodynamic conditions. At high flow rates, two distinct dissolution patterns are evidenced: for Euville samples, dissolution is distributed across the entire width of the sample, whereas for Lavoux samples, dissolution is localized in wormhole-type channels. At lower flow rates, similar wormhole-type dissolution patterns are observed for both Euville and Lavoux samples. We explain these contrasting responses by differences in pore interconnectivity, resulting in different accessible specific surface areas and reactivity. Finally, numerical transport modeling approach using Lattice-Boltzmann Method is proposed to support our hypotheses.
Laboratoire de météorologie dynamique
Date: March 13, 2026
Time: 14h
Location: salle Lien Hua – E350
By: Tom Beucler (Univ. Lausanne)
Title: From Atmospheric Physics to Machine Learning and Back
Abstract: Machine learning (ML) is revolutionizing atmospheric modeling across scales, yet ML models may violate physical laws, struggle outside their training set, and explaining their added value remains challenging—especially for deep learning models. This presentation explores a two-way synergy between ML and physical knowledge: (1) using physics to constrain or guide ML to improve its consistency and generalizability across atmospheric regimes, and (2) distilling knowledge from successful ML models via Pareto-optimal model hierarchies. I will demonstrate this with case studies, including improving the generalization of neural network parameterizations across climates, discovering equations linking cloud cover to its thermodynamic environment, and elucidating three-dimensional patterns in radiative feedbacks associated with early tropical cyclone intensification. While the focus is on weather and climate applications, the methodological frameworks apply broadly to scientific ML, with the dual purpose of improving the trustworthiness of ML for environmental applications and facilitating data-driven discovery in Earth sciences.
Laboratoire de météorologie dynamique
Date: March 12, 2026
Time: 11h
Location: salle Claude Froidevaux – E314
By: Seraphine Hauser (ETH Zürich)
Title: A quasi-Lagrangian perspective on the role of dry and moist processes for atmospheric blocking formation and maintenance
Abstract: Atmospheric blocking refers to quasi-stationary, large-scale atmospheric patterns characterized by a dominant anticyclonic anomaly that effectively “blocks” or redirects midlatitude weather systems. Despite progress in numerical weather prediction, blocking remains difficult for weather and climate models to represent due to the complex multi-scale processes involved in its lifecycle. Moreover, the persistence of these patterns frequently triggers extreme surface weather, including prolonged heatwaves, cold spells, droughts, heavy precipitation events, and associated flooding. While recent studies highlight the importance of latent heat release in building and maintaining the upper-level anticyclonic anomaly, different perspectives attribute varying roles to dry and moist dynamics. It also remains unclear whether their relative contributions differ across regions where blocking occurs.
The first part of the seminar introduces the quasi-Lagrangian potential vorticity (PV) framework, which tracks negative upper-tropospheric PV anomalies associated with blocking and quantifies the processes that govern changes in block amplitude. This approach enables an assessment of the respective roles of dry and moist dynamics. The framework has been applied to the European Centre for Medium-Range Weather Forecasts (ECMWF) ERA5 reanalysis dataset (1979–2021). In the second part of the seminar, the quasi-Lagrangian framework is applied to blocks embedded within four anticyclonically dominated weather regime types in the North Atlantic–European region: Greenland Blocking, Atlantic Ridge, European Blocking, and Scandinavian Blocking. The focus is placed on the pre-history of blocks prior to blocked regime onset and on the role of dry and moist processes in the amplification of regime-related PV anomalies. Finally, the third part presents ongoing work examining how the underlying dynamics (dry versus moist) differ among four Northern Hemisphere blocking types: Omega blocking, dipole (Rex) blocking, anticyclonic Rossby wave breaking (RWB), and cyclonic RWB.
Bureau des longitudes
Date: March 4, 2026
Time: 14h30
Location: salle Dussane – 45 rue d’Ulm et en visio ici
By: Françoise Vimeux (Institut de Recherche pour le Développement)
Title: Comment le changement climatique affecte-t-il le cycle de l’eau ?
Abstract: L’accumulation de chaleur dans le système climatique depuis l’ère préindustrielle, à cause des activités humaines, provoque une élévation de la température de surface à l’échelle planétaire qui a atteint +1,3°C au cours de la dernière décennie relativement à la seconde moitié du 19ème siècle.
L’exposé montrera en quoi cette élévation de température dans l’atmosphère et dans les océans perturbe déjà le cycle de l’eau, tant sur les réservoirs qui le constituent que sur les flux à leur interface, et comment ces modifications vont évoluer au cours du 21ème siècle selon le degré de réchauffement.
En particulier, l’exposé pendra le temps d’aborder divers aspects qui touchent aux changements attendus pour les pluies, avec une attention particulière sur les évènements de pluies extrêmes. Nous passerons aussi en revue quelques autres conséquences d’un climat plus chaud sur le cycle de l’eau comme les modifications de teneur en eau des sols, ou encore la disparition programmée des glaciers, grands réservoirs d’eau douce. Les répercussions de ces changements sur les sociétés humaines et les écosystèmes seront évoqués tout au long de l’exposé.
Laboratoire de météorologie dynamique
Date: February 25, 2026
Time: 14h
Location: salle Claude Froidevaux – E314
By: Farid Ait-Chaalal (Moody’s Analytics)
Title: Simulating Extreme Events and Their Impacts: The Science of Catastrophe Modelling
Abstract: Catastrophe models estimate financial losses from natural and man-made disasters. They are used by insurers, reinsurers, financial institutions, and public agencies for underwriting, capital management, and regulation. Historical loss records are short in comparison to the low frequency of catastrophic events. Extrapolating from history can only provide very uncertain estimates of tail risk. Catastrophe models look beyond history and fill the gaps where events could happen but have not yet occurred. To this end, cat models typically simulate thousands of years of synthetic events. I will explain how these models are built and validated, from hazard to vulnerability to financial loss. I will focus on weather and climate models and show how physical models, historical observations, and statistical methods are used to build catastrophe models. I will highlight how academic research is used and translates into operational models.
Laboratoire de Géologie
Date: February 24, 2026
Time: 11h
Location: salle Claude Froidevaux – E314
By: Souleymane Diop (LGENS)
Title: Accounting for surface albedo alters the climate benefit of conservation agriculture in Sub-Saharan Africa
Abstract: Conservation agriculture (CA) is often presented as a promising pathway for climate change mitigation. By reducing tillage and maintaining crop residues on the soil surface, CA is expected to enhance soil carbon storage and limit greenhouse gas emissions. However, its overall climate impact remains debated, because these practices also modify the land surface itself, notably by changing surface albedo the fraction of incoming solar radiation that is reflected back to the atmosphere. In this study, we explored how these biogeochemical and biogeophysical processes interact in real agricultural systems. Using two long-term experimental sites located in a subhumid tropical region and characterized by contrasting soil types, we assessed the combined climate effects of no-tillage and mulching. Our results show that no-tillage alone did not significantly affect soil organic carbon stocks, whereas mulching consistently increased carbon storage at both sites. In contrast, neither practice substantially altered nitrous oxide emissions. Beyond these biogeochemical responses, clear biogeophysical effects emerged. No-tillage systematically increased surface albedo, leading to a cooling effect regardless of soil type. Mulching, however, produced contrasting outcomes: on dark Ferralsols, surface residues increased albedo and generated a cooling effect, while on brighter Lixisols, mulching provided little cooling and even led to a warming effect when residues were maintained after harvest. When these effects were integrated and upscaled to Sub-Saharan Africa, the picture became even more nuanced. Depending on soil brightness, mulching could either contribute to net cooling up to –0.75 Tg CO₂ equivalent per year on dark Ferralsol or, conversely, result in net warming of up to +0.51 Tg CO₂ equivalent per year on bright Lixisols.
Taken together, these findings highlight that the climate benefits of conservation agriculture cannot be evaluated solely through changes in soil carbon or greenhouse gas emissions. Changes in surface albedo can either reinforce or counterbalance these benefits, depending on local soil properties. This underscores the need to systematically integrate biogeophysical effects into assessments of agricultural climate mitigation strategies.
Laboratoire de Géologie
Date: February 10, 2026
Time: 11h
Location: salle Claude Froidevaux – E314
By: Philippe Leroy (BRGM) et Alexis Maineult (LGENS)
Title: A mechanistic model for the complex conductivity of clay materials, theory and applications to experimental data
Bureau des longitudes
Date: February 4, 2026
Time: 14h30
Location: salle Dussane – 45 rue d’Ulm
By: Félix Pérosanz (CNES)
Title: Géodésie spatiale : enjeux et missions
Abstract: La géodésie spatiale est la discipline qui étudie la gravité, la cinématique et l’orientation de la Terre à l’aide de techniques spatiales. Elle fournit les systèmes de référence terrestres nécessaires à tout géo-référencement, ainsi que les paramètres permettant de les relier aux systèmes de référence célestes. Elle permet aussi avec une précision toujours meilleure de mesurer et modéliser les déformations de la surface terrestre et la redistribution des masses du système Terre dans sa globalité. La géodésie spatiale joue donc un rôle fondamental et de plus en plus central dans la compréhension des évolutions du système Terre.
Aujourd’hui, un enjeu majeur consiste à maintenir et améliorer un repère géodésique terrestre mondial avec une stabilité millimétrique sur plusieurs décennies. Cette exigence découle du besoin de mesurer avec fiabilité des phénomènes subtils mais déterminants comme la montée du niveau des mers, le rebond post-glaciaire ou les mouvements tectoniques.
Un deuxième enjeu clé concerne la mesure du champ de gravité terrestre et de ses variations temporelles. Les redistributions de masse liées à la fonte des calottes polaires, à l’évolution des nappes phréatiques, aux échanges océan-atmosphère ou aux processus internes de la Terre doivent être quantifiées avec une meilleure résolution spatiale et temporelle. La géodésie spatiale contribue également au suivi des risques naturels (séismes, glissements de terrain, volcanisme) et au développement de services de positionnement et de navigation toujours plus précis pour des usages scientifiques et sociétaux.
Les futures missions spatiales GENESIS et NGGM de l’ESA répondront en partie à ces deux enjeux et serviront d’illustration au propos.
Laboratoire de Géologie
Date: February 3, 2026
Time: 11h
Location: salle Claude Froidevaux – E314
By: Samuel Chapman (LGENS)
Title: Elastic Properties of Fluid Saturated Rocks at Seismic Frequencies
Laboratoire de météorologie dynamique
Date: January 30, 2026
Time: 14h
Location: salle Claude Froidevaux – E314
By: Antoine Doury (CNRM)
Title: Emulateurs de modèles de climat régionaux : Descente d’échelle hybride pour la production d’ensembles à haute résolution / Regional climate model emulators: Hybrid downscaling for the production of high-resolution ensembles
Abstract: Dans les dernieres années, grâce aux developements des méthodes récentes d’apprentissage statistique et notamment des reseaux de neuronnes, les emulateurs de modèles de climat régionaux (RCM) sont apparus comme une solution prometteuse voire indispensable pour correctement étudier les impacts du changement climatique aux échelles locales. L’objectif principal consiste à apprendre une relation de descente d’échelle entre une description « grande échelle » de la situation météorologique (basse résolution, moyenne/haute atmosphère) et une ou des variables de surface à haute résolution (celle du RCM) à une échéance temporelle donnée. Une fois estimée, cette fonction de descente d’échelle peut être appliquée à n’importe quelle simulation provenant d’un modèle global à basse résolution afin d’en produire une information à haute résolution. Grace au très faible coup de calcul des méthodes statistiques, de grands ensembles peuvent être créés afin d’étudier correctement le changement climatique local, c’est-à-dire en intégrant les différences sources d’incertitudes. Dans cet exposé, je présenterai et validerai le concept d’émulateur développé au CNRM, puis illustrerai un exemple d’application.
Département de Géosciences
Date: January 20, 2026
Time: 11h
Location: salle Claude Froidevaux – E314
By: Kristel Chanard (IPGP), Aglaé Jézéquel (LMD)
Title: Attaques contre l’Université et la Recherche : quelles résistances possibles ?
Laboratoire de météorologie dynamique
Date: January 15, 2026
Time: 11h
Location: salle Claude Froidevaux – E314
By: Boris Sauterey (LOCEAN)
Title: Darwinian adaptation of plankton in global ocean models
Abstract: Plankton communities are an essential component of ocean biogeochemistry and play a key role in making oceans an important climatic buffer. In the oceans, the environmental control of planktonic activity is modulated by the composition and diversity of plankton physiological traits (e.g., size, temperature and light preferences, stoichiometry, etc.). Yet, very little is known about how plankton communities assemble in the ocean under the combined influence of biological (eco-evolutionary dynamics) and physical mechanisms (mixing, transport). Moreover, this key process is very crudely represented for in current ocean models. Here, I show how integrating Darwinian adaptation into ocean models allows simulating how the functional composition and diversity of plankton communities is shaped by adaptation and ocean physics, how it feeds back on ocean biogeochemistry, and what the implications are for the resilience of marine ecosystems under climate change.
Laboratoire de Géologie
Date: January 13, 2026
Time: 11h
Location: salle Claude Froidevaux – E314
By: Alexandrine Gesret (Mines Paris – PSL)
Title: High resolution seismic imaging and uncertainty quantification: from subduction zone to reservoir
Soutenance d’HDR
Date: January 9, 2026
Time: 14h30
Location: salle salle Claude Froidevaux – E314
By: Benoit Vittecoq (BRGM)
Title: Hydrogéologie des aquifères volcaniques de Martinique et de Mayotte
Abstract: Cette soutenance d’HDR présentera une synthèse de mes travaux de recherche, à l’interface entre hydrogéologie et géophysique, pour mieux comprendre les aquifères volcaniques de Martinique et de Mayotte, des systèmes aquifères complexes, sensibles aux forçages externes (sismicité, climat, pressions anthropiques).
La première partie retracera mon parcours scientifique et professionnel, mon activité d’encadrement, ainsi que ma production académique et technique. La seconde partie présentera une synthèse de mes travaux de recherche. Cinq publications majeures font ensuite l’objet d’une analyse critique. Enfin, je proposerai dans la troisième partie un programme de recherche que je souhaite développer au cours des quatre prochaines années.
J’espère que ces connaissances pourront contribuer à un meilleur accès à l’eau pour les populations, ainsi qu’à la préservation de cette ressource vitale.
Jury: Jean-Raynal de Dreuzy (DR CNRS, ENS Rennes)
Colette Sirieix (Professeure, Université de Bordeaux)
Jean-Christophe Comte (Reader in Hydrogeology, Aberdenn University)
Florence Habets (DR CNRS, ENS-PSL)
Julio Goncalves (Professeur, Université d’Aix-Marseille)
Sophie Violette (MCF, ENS-PSL & Sorbonne Université)
Jérôme Fortin (DR CNRS, ENS-PSL)
Alain Dupuy (Professeur, INP Bordeaux – Directeur du programme scientifique EAU du BRGM)
Laboratoire de météorologie dynamique
Date: January 8, 2026
Time: 11h
Location: salle Claude Froidevaux – E314
By: Hanh Nguyen (BOM, Australia)
Title: Large-scale to local factors influencing Sumatra squalls affecting Singapore
Abstract: The climate of the Maritime Continent, including Singapore, is influenced by a wide range of tropical drivers including the El-Niño Southern Oscillation (ENSO), the Indian Ocean Dipole (IOD), the Madden–Julian Oscillation (MJO), and equatorial wave activity. In Singapore, the rainfall pattern is often marked by episodes of short intense rainfall bursts, primarily driven by mesoscale convective systems, such as Sumatra squall lines. These episodes often result in local flash floods and strong wind bursts. Here we analyse 33 years of Sumatra squalls affecting Singapore and investigate the potential impact of the main modes of climate variability and atmospheric equatorial tropical waves. Highlighted results are that, on average, Sumatra squalls originate from Sumatra and the Strait of Malacca, make landfall over Singapore primarily in the morning and last about 2 hours, and are more frequent from April to November. These squalls tend to be more frequent under La Niña conditions, which are associated with locally warm sea surface temperature anomalies and favourable atmospheric anomalies over the Maritime Continent region. On intraseasonal scale, the MJO may play a role in setting the westerlies and enhancing convective conditions favorable for the squalls to intensify and propagate eastward. Convectively-coupled Kelvin waves are important for the initiation and eastward propagation of the squalls over Sumatra and the Strait of Malacca, while equatorial Rossby, and mixed Rossby gravity waves influence their meridional propagation toward Singapore.
Bureau des longitudes
Date: January 7, 2026
Time: 14h30
Location: salle Dussane – 45 rue d’Ulm
By: Pierre-Olivier Lagage (CEA)
Title: Mission spatiale James Webb : morceaux choisis
Abstract: Lancé le 25 décembre 2021 par une fusée Ariane, le télescope spatial James Webb révolutionne à grande vitesse notre compréhension de l’Univers. Cet observatoire, le plus ambitieux et le plus complexe jamais envoyé dans l’espace, nous ouvre l’accès à des régions jusqu’ici totalement inexplorées. Grâce à son immense miroir et à sa vision en lumière infrarouge, le James Webb permet de remonter à l’époque dite de la sortie de « l’âge sombre » de l’Univers, lorsque se sont formées les premières étoiles, les premières galaxies et les premiers trous noirs. Il offre également une vision sans précédent des nuages de gaz et de poussière où naissent les étoiles et leurs systèmes planétaires, et permet pour la première fois d’étudier en détail les atmosphères d’exoplanètes. Ainsi, presque tous les domaines de l’astrophysique sont aujourd’hui profondément transformés par ses observations.
Après une brève présentation du télescope James Webb et des innovations technologiques qui ont rendu cette mission possible, j’illustrerai quelques propriétés de la lumière infrarouge à l’aide d’une petite caméra infrarouge. Je présenterai ensuite plusieurs résultats marquants, en mettant particulièrement l’accent sur les découvertes récentes concernant les exoplanètes.