Atelier national sur les nuages polaires

24/06/2025 09:00

SIRTA / ICEO : Journée Scientifique 2025

24/06/2025 09:00

Evénement de clôture projet FAIR-EASE

12/06/2025 09:00

Les campagnes d’opportunité : des partenariats éthiques pour la recherche scientifique ?

19/12/2023 11:00

Séminaire du LOCEAN.

Sciences sociales et adaptation: l'enjeu des résistances et des moteurs à l'élaboration des stratégies territoriales d'adaptation

15/12/2023 11:00

Webinaire TRACCS.

Upper Ocean Nutrients Budget and Biological Pump Response to the Rapid Sea Ice Retreat

15/12/2023 11:00

Séminaire du LOCEAN.

Relationship and feedback between LULC changes and hydroclimatic variability in Amazonia

06/05/2024 14:00

The Amazon rainforest plays a vital role by functioning as a regulator of the climate system, providing essential ecosystem services and acting as the main terrestrial carbon sink. It drives hydroclimatic processes and mitigates the effects of droughts through the strong coupling exerted between the soil, vegetation and the atmosphere. Indeed, forests operate as hydraulic pumps, absorbing and linking water stored in the soil with the atmosphere. Therefore, they have the potential to impact rainfall patterns through biophysical processes like water recycling. However, these capacities have been reduced during the last decades due to disturbances in the climate-vegetation system. The hydrological cycle has intensified, and extensive forest areas have been degraded. All this has accentuated a process of biophysical transition from a predominantly forested ecosystem to a Savanna. Therefore, given these complexities, understanding the direction of changes, the impacts and their interrelationships is of vital importance.

Using multiple datasets and a Land Surface Model coupled with a General Circulation Model, this thesis delves into the study of the interactions between Amazon hydroclimatology and vegetation. In addition, it seeks to expand our understanding of modifications in the vegetation-atmosphere system and its links with climate and Land Use and Land Cover (LULC) changes. Likewise, taking into account the increasing rates of deforestation, it investigates the effects and feedback resulting from a large-scale forest loss scenario on hydrological processes and water stress.
The results show that, over the southwestern Amazon, a little-explored region, forests undergo a transition from being influenced by energy availability to depending on water availability throughout the year. During the peak of the rainy season, vegetation growth is primarily influenced by energy availability rather than water availability. Nevertheless, outside of this period and for most of the year, forests respond positively to precipitation and terrestrial water storage, suggesting that vegetation is primarily dependent on water supply. However, a spatial analysis reveals that recent deforestation modifies these transitions and destabilizes the natural balance in the climate-vegetation system.

The nature of these imbalances in the Amazon is not entirely clarified. Through an approach based on the relationships of water fluxes, energy and vegetation conditions over the last four decades, it is explored whether these changes are intrinsic to climate variability or are driven by anthropogenic processes. 67% of the southwestern Amazon has experienced a transition towards a predominantly dry state due to climatic factors (external forcing), while 21% has transitioned towards a state dominated by deforestation (internal forcing). However, external and internal forcings are not independent processes, as both mechanisms drive changes simultaneously. By weighing the magnitudes of these forcings, we show that the synergies have led 74% of the southwestern Amazon toward a state of greater water stress. Nevertheless, during recent years, although combined external-internal processes continue to exert significant control over changes, 30% of these are strictly dominated by internal forcing. This suggests that internal processes are playing an increasingly relevant role in the transition towards a state characterized by high forest water stress, especially in areas where deforestation and anthropogenic pressure are increasing.

Using the land surface component (ORCHIDEE) coupled to the atmospheric component (LMDZ) of the IPSL Earth System Model, the effects of projected deforestation by 2050 on the terrestrial hydrological cycle and dryness of the Amazon region are examined. Deforestation decreases precipitation, reduces evapotranspiration and increases runoff. In addition, it drives an increase in the runoff coefficient and a significant decrease in moisture recycling, which results in a water deficit throughout the Amazon ecosystem. In fact, deforestation accentuates water stress especially in the southwestern Amazon (positive feedback). Water demands in the atmosphere, on the land surface and even in the soil root zone intensify during the dry season. During the wet season, the deficit of specific atmospheric humidity becomes even more acute towards the tropical Andes over the Altiplano region. These findings provide a more thorough understanding of the possible effects of massive forest removal on the water availability and resilience of the Amazon in a context where changes are occurring at an accelerated rate.

Sea-level rise and coastal risks: marine flooding and coastal erosion

23/05/2024 14:00

Coastal hazards such as flooding, erosion and salinization are a major concern for coastal zones management. In fact, projections suggest that by the middle of the 21st century, approximately 1 billion persons will be exposed to these hazards globally due to ongoing sea-level rise and coastal development. Confronted with this challenge, coastal managers and researchers are asking the same questions: (1) are we already able to attribute specific coastal impacts to sea-level rise? (2) how, where and when should climate change and sea-level rise impacts materialize? Responding to these two questions is generally difficult due to the limited accuracy of coastal hazard and risk models. This leads to a third question: how can we evaluate and manage uncertainties in sea-level rise and coastal impact projections?

Within my research, together with many colleagues, I adapted, developed and applied methods to address these three questions. This includes approaches to detect and eventually attribute impacts of sea-level rise as well as probabilistic methods to propagate uncertainties from coastal forcing (such as sea levels, waves and surges) to coastal impacts such as flooding and erosion. Yet, probabilistic approaches have a limited ability to capture future coastal risks because the probability of an early ice-sheet collapse during the late 21^st century or early 22^nd century is unknown. To model this deep uncertainty, we proposed to go beyond the use of single probabilistic distributions and use extraprobabilitic approaches to represent future sea-level changes and propagate them across coastal impacts models and eventually support some coastal adaptation decisions.

Over the coming years, it is clear that present-day and future coastal risk assessments will become more precise, or at least that the assumptions of these models will be clearly set out. For example, the CoCliCo project, which I am coordinating, aims at developing such broad-scale projections of coastal flooding in Europe. However, by limiting ourselves to delivering information on future risks without assessing critically adaptation options, we may be missing the most important aspect of adaptation. Indeed, the challenge in coastal zones is not limited to protecting against coastal hazards and sea-level rise. It rather consists in achieving what the IPCC calls climate resilient development, that is, mitigating climate change, adapting to committed impacts of climate change, reducing biodiversity losses and achieving the 17 sustainable development goals adopted by United Nation members in 2015. As part of my future research projects, I propose to contribute exploring pathways toward coastal resilient development.

Restitution synergique radar-lidar multiplateforme pour nuages liquides et de phase mixte

05/04/2024 14:00

Restitution synergique radar-lidar multiplateforme pour nuages liquides et de phase mixte

Les nuages jouent un rôle important dans le cycle de l’eau et le bilan radiatif de la Terre, et tendent à légèrement refroidir le climat. Cependant, de nombreuses incertitudes demeurent concernant leurs rétroactions et leur évolution dans le contexte du réchauffement climatique. Les nuages de phase mixte représentent notamment une part significative de l’effet radiatif des nuages. Ils sont constitués d’un mélange de cristaux de glace, de gouttelettes d’eau surfondues et de vapeur d’eau. Cette coexistence implique des processus complexes et la fraction de liquide et de glace affecte de manière significative leurs propriétés radiatives. Cette complexité les rend difficiles à représenter dans les modèles numériques, introduisant des biais significatifs. Il est donc crucial de mieux comprendre les processus microphysiques de ces nuages pour réduire les incertitudes des prévisions climatiques et météorologiques.

Pour observer les nuages, il existe plusieurs types d’instruments, tels que les sondes in situ (directement au contact des hydrométéores) et les instruments de télédétection (observations distantes). Les radars et les lidars nous permettent d’obtenir des informations résolues en distance et peuvent être embarqués à bord d’avions ou de satellites, offrant ainsi couvertures régionale et globale. Les radars nuages travaillent à des fréquences (35 et 95 GHz) auxquelles la réflectivité est sensible à la taille des particules, impliquant une réflectivité plus élevée pour les grosses particules nuageuses (les cristaux de glace) que pour les petites particules (les gouttelettes d’eau).

Les lidars, quant à eux, fonctionnent habituellement entre 355 et 1064 nm et sont globalement plus sensibles à la concentration des particules. Ainsi, la rétrodiffusion lidar est plus élevée pour les particules très concentrées, telles que les gouttelettes d’eau. Leur synergie permet de tirer avantage des forces et des faiblesses de chacun pour restituer les propriétés des nuages. Cependant, ces propriétés ne sont pas directement accessibles à partir des mesures et des algorithmes de restitution sont donc utilisés pour relier les mesures aux propriétés microphysiques.

Cette thèse propose une nouvelle méthode synergique radar-lidar dédiée à la restitution des propriétés des nuages d’eau surfondus, de glace et de phase mixte. Sur la base d’une méthode existante mais dédiée uniquement aux nuages de glace, une nouvelle approche permettant d’inclure à la fois l’eau surfondue et les situations de phase mixte a été développée.

La première étape a été d’adapter et d’améliorer la classification servant à identifier la nature des particules observées. Ensuite, de nombreuses adaptations ont été apportées à l’algorithme afin de restituer séparément les propriétés des cristaux de glace et de l’eau surfondue. Cette approche est basée sur les sensibilités différentes du radar et du lidar vis-à-vis des deux types d’hydrométéores : les cristaux de glace dominent le signal radar tandis que l’eau surfondue domine le signal lidar. Afin d’évaluer cette nouvelle méthode, les restitutions sont comparées à des mesures in situ, provenant d’observations colocalisées et de la littérature. La première étude compare les restitutions obtenues à partir des données satellites CloudSat-CALIPSO avec des mesures in situ aéroportées colocalisées.

Cette étude montre que les restitutions radar-lidar suivent les mêmes tendances que les mesures in situ et fournissent des résultats prometteurs avec un pourcentage d’erreur moyen de 49 % pour le contenu en eau liquide et 75 % pour le contenu en glace et ce malgré des échelles de mesures différentes et une colocalisation imparfaite. La méthode développée est également appliquée aux plateformes aéroportées française et allemande RALI et HALO.

Les premiers résultats sont prometteurs et les données in situ colocalisées obtenues lors de campagnes récentes pourront être utilisées pour évaluer davantage l’algorithme et améliorer son paramétrage.

 

---

Multiplatform radar-lidar synergistic retrieval for liquid and mixed-phase clouds

Clouds play an important role in the Earth’s water cycle and radiation balance, and tend to cool the climate slightly. However, there are still many uncertainties about their feedbacks and their evolution in the context of global warming. In particular, mixed-phase clouds account for a significant proportion of the cloud radiative effect. They are composed of a mixture of ice crystals, supercooled water droplets and water vapor. This coexistence involves complex processes and the fraction of liquid and ice significantly affects their radiative properties. This complexity makes them difficult to represent in numerical models, which introduces significant biases. For this reason, it is crucial to better understand the microphysical processes of these clouds to reduce the uncertainties in climate and weather forecasts.

To observe clouds, several instrument types exist, such as in situ probes (in direct contact with the hydrometeors) and remote sensing instruments (remote observations). Radar and lidar allow us to obtain distance-resolved information. They can be deployed onboard aircraft or satellites, providing regional and global coverage. Cloud radars work at frequencies (35 and 95 GHz) at which the reflectivity is sensitive to particle size, implying higher reflectivity for large cloud particles (ice crystals) than for small particles (water droplets). Lidars, on the other hand, usually operate between 355 and 1064 nm and are generally more sensitive to particle concentration. As a result, lidar backscatter is higher for highly concentrated particles, such as water droplets. Their synergy allows us to take advantage of the strengths and weaknesses of each instrument to retrieve cloud properties. However, these properties are not directly accessible from measurements and retrieval algorithms are therefore used to relate measurements to microphysical properties.

This thesis proposes a new radar-lidar synergistic method dedicated to retrieve supercooled water, ice and mixed-phase cloud properties. Based on an existing method dedicated solely to ice clouds, a new approach has been developed to include both supercooled water and mixed-phase situations. The first step was to adapt and improve the classification used to identify the nature of the observed particles. Next, numerous adaptations have been applied to the algorithm to retrieve separately ice crystals and supercooled water properties. This approach is based on the different sensitivities of radar and lidar to the two types of hydrometeors: ice crystals dominate the radar signal while supercooled water dominates the lidar signal.

To assess this new method, the retrievals are compared to in situ measurements from co-located observations and the literature. The first study compares retrievals from CloudSat-CALIPSO satellite data with collocated in situ airborne measurements. This comparison shows that the radar-lidar retrievals follow the same trend as the in situ measurements and provide promising results with mean percent error of 49 % for liquid water content and 75 % for ice water content, despite the quite different measurement scales and imperfect collocation. Additionally, this has been applied to the French and German airborne platforms RALI and HALO. These first results are promising and the collocated in situ data collected during recent campaigns can be used to further assess the algorithm and improve its parameterization.