As climate change continues, high quality climate information over the short to medium term (years to decades) is required in order to foresee, mitigate, and adapt to its impacts. Currently, much of this information is created using climate models of high complexity. As such, there are two key challenges in climate science: 1) Creating climate models that are of high quality; for example, balancing the improvements from increased complexity with increased computational expense and the exploration of potential sources of parameter uncertainty, and 2) Effectively exploiting the outputs of multiple models; for example, the projected 18 PB of data representing approximately 100’000 total simulated years on new multi-model archives (“Big Data”). In HARMONY, we propose a combined strategy to harness this data to both efficiently improve climate models and make better predictions and projections of future climate change over Europe. We are now at an inflection point, where these efficiencies and improvements can be made a reality. A combination of state-of-the-art new approaches in Machine Learning (in particular in interpretability) and Uncertainty Quantification are themselves made possible by new infrastructure: the vast, new 6th Coupled Model Inter-comparison Project (CMIP6) archive and a game-changing new GPU cluster at the Host Institution dedicated to Machine Learning. In HARMONY, we will integrate these new approaches and infrastructure to efficiently find and exploit the high quality - but often hard to access - information within climate models. The primary outcomes of HARMONY will be threefold: 1) Development of new Machine Learning approaches and Interpretability tools, resulting in greater understanding of European climate variability, leading to 2) A demonstrable framework for the efficient improvement of climate models, yielding more robust climate projections, and 3) Increased skill in - and understanding of - near term predictions of European climate.

The scope of the OMAGE project is to better understand the dynamics of giant planet's atmospheres as a whole, by associating remote sensing observations to the development of a novel, state-of-the-art, General Circulation Model (GCM). Recent observational programs, both spatial and ground-based, have revealed the complexity of giant planet's middle atmospheres. In particular, for Saturn, maps of the temperature and of the distribution of trace species have been obtained by the Cassini spacecraft with unprecedented details. These maps exhibit puzzling anomalies which cannot be explained by current photochemical and radiative models (none of them includes dynamics), and which have been interpreted as signatures of large-scale or seasonal dynamical motions. However, as no model of Saturn's stratospheric circulation currently exists, these assumptions have not yet been tested, and Saturn’s global circulation remains weakly characterized. Furthermore, on Saturn and Jupiter, equatorial oscillations in the zonal wind and temperature field have recently been discovered and are reminiscent of the Earth's Quasi-Biennial Oscillation, a fundamental dynamical phenomenon. These oscillations thus appear to be a common dynamical phenomenon in very different planetary atmospheres. In this project, we propose to study in detail the atmospheric circulation of giant planets by developing the first general circulation model of their stratospheres. It will serve as a new tool to address fundamental questions in geophysical fluid dynamics, explore the giant planets circulations and better interpret current and future observations. The development of the GCM for gas giants will be based on 1) our current knowledge of their physical characteristics, brought forward by recent, high-quality observations and 2) the expertise of the Laboratoire de Météorologie Dynamique (LMD) for developing GCMs of the atmospheres of Venus, Mars and Titan over the past two decades. This new GCM will first be focused on reproducing Saturn's climate, for which observational constraints are the more documented. Then, we will move forward to study Jupiter, which presents similar problematics. Several case studies of comparative planetology will be investigated. Finally, this model will be adapted to extrasolar planets such as « hot Jupiters » that act as natural laboratories for this new tool, to broaden our knowledge of atmospheric dynamics in extreme environments. The first part of this project is focused on adapting the current LMDz GCM to Saturn. Several tasks are defined and include: - Adapting the existing dynamical core to gas giants, - Adapting the physics packages (radiative transfer, chemistry...) to Saturn's specific conditions (radiative times, composition...), - Parameterizing boundary conditions of these gas planets, which are very different from terrestrial planets, - Optimizing the efficiency of calculations, as this new model will be costly in computation time. In parallel, we will contribute to the analysis of new Cassini data (in orbit around Saturn until 2017), in order to monitor seasonal variations in temperature and composition, that will bring mandatory constraints to the model. When the GCM is mature enough, we will investigate the steady state circulation that takes place on the planet. Our objectives will be: - Characterizing the global and/or seasonal circulation cells, - What is their efficiency to transport trace species and can we reproduce Saturn's observed distribution of hydrocarbons, - Studying the role of the ring's shadows (that moves from one hemisphere to another with seasons) on seasonal forcing, - Studying what triggers and governs the evolution of the equatorial oscillation, - Explore wave activity and their role in the observed phenomena. In the last part of the project, we will apply Saturn's GCM to other gas giants : first Jupiter, quite similar to Saturn, then extrasolar planets, characterized by their extreme conditions.

The effect of climate change and oceanic carbon uptake on marine biogeochemistry remains highly uncertain. In particular, we have limited understanding of how the temporal variability of ocean chemistry will change over the 21st century while impacts on biological communities are even more uncertain (Bopp et al., 2013). This project (CONVINCE) aims to significantly reduce uncertainties associated with the interactions between climate, the oceans and marine ecosystems using a combination of novel experimental and model techniques. Adopting the recent concept of Emergent Constraints, CONVINCE will explore Earth System Model (ESM) ensembles for innovative ways to constrain projections of ocean-climate interactions. Using geochemical measurements, the project will assess the natural sensitivity of European coastal ecosystems to present ocean chemistry variability. Changes in the community calcification rates of these ecosystems since the pre-industrial will be estimated using alkalinity manipulation techniques (Albright et al., 2016). The project will deliver: (i) an enhanced understanding of how the temporal variability of ocean chemistry responds to climate change; (ii) Emergent Constraints on interactions between climate and the oceans based on satellite and shipboard observations and (iii) an in situ assessment of the impact of ocean acidification since the pre-industrial on European coastal ecosystems. CONVINCE will benefit from the expertise and experience of the PI, which includes training as an Earth System scientist, the development of the first Emergent Constraint on climate projections of the marine realm (Kwiatkowski et al., 2017) and involvement in pioneering field experiments that aim to understand the historical legacy of ocean acidification (Albright et al., 2016) and the importance of temporal ocean acidification variability for marine communities (Kwiatkowski et al., 2016a).

EMERGIANT is a research project about Jupiter and Saturn, which lies at the interface between astrophysics and atmospheric science. EMERGIANT is fundamental research, driven by the incentive to learn more about the world we live in (the overarching goal of the ANR 2017 Défi des autres savoirs). The extreme weather of Jupiter and Saturn, our Solar System's so-called gas giants, is a permanent source of wonder and inspiring challenges. The community has been debatting for decades the mechanisms underlying fast-rotating gas giants' spectacular banded jets (strong planetary-scale wind currents), long-lasting traveling storms (e.g. centuries-old Jupiter's Great Red Spot), and powerful convective thunderstorms (e.g. Saturn's Great White Spots). Enigmatic differences between Jupiter and Saturn are still yet to be explained: although Saturn receives less sunlight and generates less internal heat than Jupiter, its equatorial super-rotating jet is three times stronger; and contrary to Jupiter, Saturn exhibits a remarkable hexagonal-shaped circumpolar jet. We are currently living an observational golden age for gas giants, with a wealth of observations on a variety of atmospheric phenomena in Jupiter and Saturn's troposphere and stratosphere -- raising new constraints and questions to achieve the overarching goal of fully understanding Jupiter and Saturn's atmosphere and climate. We propose with the EMERGIANT project to perform numerical atmospheric modeling on French supercomputers to make a paradigm shift from the phenomenological point of view imposed by observations, to a complete physically-based understanding of gas giants' atmospheres. We will carry out unprecedented dynamical simulations with a full Global Climate Model for gas giants to draw the big picture of gas giants' climate and meteorology. Not only this is a unique methodology to draw a comparative planetology approach between Jupiter and Saturn, but this will open broader perspectives of geophysical fluid dynamics' questions also valid on the Earth: the driving of jets by eddies from an inverse energy cascade, the emergence of cyclones / anticyclones, the impact of moist convection on the large-scale planetary circulations, and the stratospheric oscillations reminiscent of the Quasi-Biennal Oscillation on Earth. EMERGIANT will allow the scientific coordinator to build a strong local team at Laboratoire de Météorologie Dynamique / Université Pierre et Marie Curie, which will evolve in a high-level network of national, European, and international collaborations. The EMERGIANT project will enhance the scientific return of European and international space missions, as well as High Performance Computing capabilities in France and Europe. The simulation results obtained with the EMERGIANT gas giants Global Climate Model will be made available through an online graphical interface opened to scientific communities, students and the general public.

Despite their occurrence in the upper troposphere, ice-cloud formation remains one of the least understood processes in the atmosphere and one of the largest uncertainties in the prediction of climate change (IPCC 2014). However, coupling between in-situ aerosol measurements and computational models continue to improve our understanding of the formation and radiative effects of clouds in the atmosphere (Flossman and Wobrock, 2010). The principal objective of this research project is to complete the LaMP instrumental platform, through construction of an ice nucleation (IN) chamber that will be capable of sampling ambient and artificial aerosol particles into environments which permit the formation of ice crystals. The second objective of this project is to perform long-term in-situ measurements to try to understand the meteorological conditions and physical and chemical characteristics of aerosol particles required to act as ice nuclei. Once we understand the physical and chemical properties required to form an ice crystal we can begin to predict their formation in the atmosphere. The principal objectives of the laboratoire de Meteorologie physique are to study atmospheric particles and gases, and their chemical and physical interactions with cloud droplets. These research objectives are achieved through a combination of in-situ measurements and modelling studies. Our studies on clouds are focused on both aqueous-phase droplets and ice-crystals with in-situ measurements being made aboard an airborne platform and at the puy de Dome research station. Currently the LaMP is considered the leading French laboratory in ice-cloud experimental research as well as being recognised internationally for its ability to model liquid, mixed-phase, and ice clouds. This proposal is focused on understanding ice crystal formation in the atmosphere, and to characterise the chemical and physical properties of ice crystal residues. • The principal objective of this research project to complete the instrumental platform, through construction of an ice nucleation (IN) chamber that will, firstly, be capable of sampling ambient aerosol particles into environments which permit the formation of ice crystals. Secondly, it will be capable of introducing artificial aerosol particles into the chamber so that the dependence of the ice nucleating ability on aerosol composition, morphology, and concentration can be determined in different atmospheric environments (polluted vs. background). Thirdly this chamber will have a unique feature where it will be possible to chemically and physically characterise the ice crystals residues exiting the chamber. This feature is currently not performed with existing IN chambers. • The second objective of this project is to perform long-term in-situ measurements to try to understand the meteorological conditions and physical and chemical characteristics of aerosol particles required to act as ice nuclei. We plan to achieve this through coupling the IN chamber using a combination of online instruments (aerosol mass spectrometer, scanning particle mobility sizers) and offline instruments such as electron microscopy at the puy de Dome research site. This combination of instruments will provide us with detailed information on aerosol chemical and physical properties as well as obtaining more detailed information on particle mixing state and morphology through electron microscopy measurements. The combination of the IN chamber and the online/offline measurements will allow us to characterise the chemical and physical properties of ice crystal residues. Once we understand the physical and chemical properties required to form an ice crystal we can begin to predict their formation in the atmosphere. Additionally with this information we can begin to quantify the anthropogenic effects on ice-cloud formation.