Workshop Agenda & Sessions

[1] AGENDA

On-site participants will be able to participate in all sessions. Due to logistical constraints, remote participants will only have access to Plenary discussions, short talks, and keynote talks via this link: https://videos.univ-grenoble-alpes.fr/live/events/. Registered participants will receive ZOOM links before the event in order to participate more interactively. 

The agenda below will be complemented with the list of short talks, and may be slightly updated until the workshop. The 5 types of sessions and their content are presented in detail hereafter.

[2] ON-SITE ACTIVITY : days 2,3 (first sessions)

A two-stage Hands-on session will be held on Tuesday and Wednesday mornings: during registration, please indicate if you wish to participate, and prepare your computing setup as described just below. 

NEMO/AI hybrid modelling: technical setup on Tuesday, interactive demonstration on Wednesday (Alexis Barge & J. Le Sommer)

Description: ...

Prerequisites: ...

[3] PLENARY DISCUSSIONS : days 1, 2, 3 (10:45 - 11:30am)

These sessions will be organized and guided by discussion chairs, who will summarize the discussions in the workshop closing session. Themes and schedules will be added to the website.

MONDAY : How to explain, possibly mitigate, and live with model drift ? (T. Penduff, S. Griffies)

Most ocean and climate model simulations using unassimilated primitive equation models exhibit spurious trends, also called numerical drifts: simulated oceans often tend towards a state that differs from that of the real world, on timescales ranging from a few years for small or shallow domains to several centuries for global simulations. These spurious trends not only bias the final state of simulations: they can hinder or compromise numerical studies of low-frequency variability and geophysical trends (associated, for example, with climate change) observed in the real ocean. Model drifts may potentially come from unbalanced surface or lateral forcings, numerical inconsistencies, incompatibilities between components of model configurations (initial conditions, bathymetry, forcings, parameterizations, …), etc. However, this long-standing problem is not well documented in the ocean and climate modelling literature; we propose to discuss the possible origins and consequences of these spurious trends, accepted practices for reducing them (online or during post-processing) when possible, or for living with them (sometimes at the cost of excluding certain scientific investigations...).

TUESDAY : NEMO Developments — 2026 Work Plan, ENMASSE project (P. Mathiot, S. Masson, D. Iovino) 

In this session, representatives of the NEMO System Team and the NEMO Developer Committee will provide an update on ongoing developments and future plans for the NEMO modelling framework, followed by an open discussion: 

• NEMO: Developments from the ENMASSE Project: Following the release of NEMO 5, the NEMO development team is now engaged in the ENMASSE project, which aims to deliver the next-generation NEMO modelling infrastructure. This session will present the project’s key milestones, its objectives, and the implementation timeline. 
• 2026 Work Plan Highlights: Each year, the NEMO System Team and Working Group leaders (HPC, Sea Ice, Kernel, Top, etc.) collaborate to define the community work plan. The 2026 plan will continue advancing NEMO’s scientific and technical capabilities in line with the long-term development strategy. The main priorities and planned actions for 2026 will be presented and discussed.
• Open Q&A: An open discussion will follow, providing participants with the opportunity to engage with members of the System Team and Working Groups, ask questions, and share feedback on NEMO’s ongoing and future developments.

WEDNESDAY : A NEMO analysis toolbox (S. Griffies, C. Dufour) --> Please fill out this form as soon as possible so that we can better prepare for this session : https://framaforms.org/nemo-diagnostic-tools-session-drakkar-meeting-1765549220

This plenary session aims to discuss strategies for realizing an analysis code environment/toolbox for NEMO simulations. The toolbox is envisioned as a multi-faceted online (e.g., exact budget terms) and offline (e.g., manipulations of budget terms via Python notebooks) suite of diagnostics that facilitates scientific analysis and publications. Examples include a recent project developing water mass transformation diagnostics for MOM6 simulations that can be used as a starting point for NEMO. This experience suggests that a successful analysis toolbox is optimally realized via close collaborations between code developers (those focused on the dynamical core, parameterizations, optimization) and code users (students, postdocs, researchers), with a nonzero overlap between these groups. Results from a survey of the broader NEMO community will be summarized concerning diagnostics presently available for NEMO (both online and offline), as well as those that could be included on a prioritized wish list. The aim is to offer seeds for an organized and vigorous effort to realize a community of active users and developers of NEMO analysis code and use cases.

[4] POSTER & SHORT TALKS SESSIONS : days 1, 2, 3

Poster format: A0 in portrait orientation.  

• Poster sessions form the core of the DRAKKAR workshop, and will take place every afternoon for up to 90 minutes. The daily list of posters will be given in the dedicated tab on the left. We kindly ask on-site speakers to bring and display their physical poster in the morning of the day when they will give their short talk, and to remove it on the same evening. (Posters from remote speakers may be displayed too, if they have been brought by colleagues to Grenoble).
• Short talks + Q&A sessions. These sessions include several ~5-min talks followed by a 20-minute Q&A slot. They are intended to introduce the poster sessions that takes place afterwards.

Registered participants will also have access to the PDF versions of all accepted posters about one week before the workshop. We will thus ask by mid January on-site and remote speakers to upload 2 documents (a few introductory slides for short-talk sessions, and the PDF version of their poster). On-site participants will display their physical poster in the venue for one day; remote participants will have their PDF poster available to all registered participants for virtual Q&A.

[5] KEYNOTE SPEAKERS AND TALKS : days 1, 2, 3

Three young scientists will present their recently published work during 25-30min keynote talks followed by 20 min of questions&answers (Q&A).

MONDAY : Neil Fraser (Scottish Association for Marine Science, UK). Vertical velocity dynamics in the North Atlantic and Implications for AMOC. The Atlantic meridional overturning circulation (MOC) is traditionally monitored in terms of zonally integrated transport either in depth space or in density space. The near-equivalence of the depth- and density-space MOC in the subtropics suggests that vertical and diapycnal volumes transports are intimately coupled, whereas the divergence of these two metrics at higher latitudes indicates that any such coupling is neither instantaneous nor local. Previous work has characterized the surface buoyancy forcing and mixing processes which drive diapycnal volume transport. Here, we develop a new analytical decomposition of vertical volume transport based on the vorticity budget. This analysis highlights the roles of 1) relative vorticity advection for the sinking of overflow water at the northern subpolar North Atlantic boundaries and 2) the geostrophic β effect for the sinking of dense waters in the intergyre region. We find that vertical transport associated with surface Ekman pumping dominates observed MOC variability, and that the resulting adiabatic isopycnal heave largely explains the lack of meridional coherence in the observed AMOC. Finally, we explore how using an alternative vertical coordinate which accounts for barotropic vortex stretching can provide further insights into the coupling between density- and depth-space MOC metrics.

TUESDAY : René Schubert (GEOMAR, Germany). The ocean flows downhill near the seafloor and recirculates upward above. The ocean’s circulation redistributes heat, salt, biota, dissolved gases, microplastics, and sediments on Earth. The abyssal ocean, in the lowest 1000 m above the seafloor, moves on average with the deeper seafloor to its left in the Northern Hemisphere and to its right in the Southern Hemisphere. This finding has received little attention and its consequences for the abyssal vertical circulation have remained largely unexplored. Here, we show, using current-meter measurements and numerical simulations, that the interior flow, O(100 m) - O(1000 m) above the seafloor, is deflected within the bottom boundary layer, the lowest O(10 m), into a widespread downhill flow. This flow intensifies with the steepness of the seafloor. We further reveal that typical local changes in seafloor steepness lead to a shallow divergence and a deep convergence of this downhill flow. These are connected by an overlying upward recirculation forming closed overturning cells that extend on average over the lowest 1000 m of the ocean. Finally, the effects on the abyssal buoyancy budget are discussed. As oceanic overturning is climate-relevant, our study highlights the need to better understand the associated dynamics, to identify its impacts, and to implement these impacts into climate and Earth system simulations. Furthermore, the effects of widespread near-bottom downwelling on sediments, microplastics, chemicals and biota need to be investigated in the future. doi.org/10.1038/s41467-025-61027-2 

WEDNESDAY: Jiarong Wu (Courant Institute of Mathematical Sciences, USA). ...