Projection des changements climatiques futurs

Oliviér Boucher, Jean‐Louis Dufresne, Jessica Vial, Éric Brun, Julien Cattiaux, Fabrice Chauvin, David Salas y Mélia, Aurore Voldoire, Laurent Bopp, Pascale Braconnot, Philippe Ciais, Pascal Yiou, Éric Guilyardi, Emmanuel Mignot, Céline Guivarch2015 · DOI

TL;DR

This paper presents projections of future climate change based on the CMIP5 intercomparison project, focusing on the impacts of human activities on temperature, precipitation, cryosphere, extreme weather, and the carbon cycle. While highlighting robust findings on the magnitude and geographical patterns of expected changes, it also underscores the inherent limitations and persistent uncertainties in climate modeling.

What Problem Does This Paper Address?

The paper addresses the fundamental challenge of predicting how Earth's climate system will evolve in the 21st century under the influence of human activities. This involves understanding the complex interactions within the climate system and how they respond to various anthropogenic forcings, such as greenhouse gas emissions. The authors aim to provide a clearer picture of future climate by leveraging advanced climate models and exploring different socio-economic scenarios, acknowledging that precise predictions are impossible but that robust projections can inform policy and adaptation strategies.

Methods

The study utilizes numerical simulations from climate models participating in the Climate Model Intercomparison Project phase 5 (CMIP5). A key aspect of the methodology involves exploring different socio-economic scenarios, specifically the Representative Concentration Pathways (RCPs), which define future greenhouse gas concentrations. The authors focus on illustrative results from two French climate models (CNRM-CM5.1 and IPSL-CM5A-LR) within the broader CMIP5 ensemble. They analyze the projected changes in various climate variables, including surface temperature, precipitation, cryosphere (snow cover, sea ice, glaciers), extreme weather events, and the carbon cycle. The paper also touches upon decadal prediction methods, which bridge the gap between weather forecasting and long-term climate projections.

Datasets

The paper relies on data generated from climate model simulations within the CMIP5 project. For observational data, it references HadCRUT4 for surface temperature and CRUTEM4 for European summer temperatures. The study also implicitly uses historical climate data for model evaluation and comparison.

Key Insights and Ideas

  • The future climate is not a simple extrapolation of the past: Climate models demonstrate that the climate system responds dynamically to perturbations, making direct historical analogies insufficient for future projections.
  • Socio-economic scenarios are crucial drivers of future climate: Different pathways of human development and emissions lead to vastly different climate outcomes, highlighting the importance of policy choices.
  • Robust signals emerge despite model uncertainties: While there are significant uncertainties in the precise magnitude and regional details of climate change, key trends like amplified warming over continents and in the Arctic are consistently projected.
  • Extreme events are a major concern: The paper emphasizes that beyond average changes, the frequency and intensity of extreme weather events (heatwaves, heavy precipitation) are projected to increase significantly, posing substantial risks.
  • The carbon cycle acts as a feedback loop: The way the ocean and land absorb CO2 influences atmospheric concentrations and, consequently, the rate of warming, creating a complex feedback mechanism.
  • Climate modeling is an evolving science: The study acknowledges that despite advancements, persistent uncertainties remain, particularly concerning cloud feedbacks and regional climate responses.

Key Results

  • Global Temperature: Projections show a significant warming trend throughout the 21st century, with the magnitude strongly dependent on the emission scenario (e.g., RCP8.5 leading to much higher warming than RCP2.6).
  • Regional Temperature: Warming is projected to be amplified over continents and at high latitudes in the Northern Hemisphere.
  • Precipitation: While global precipitation is expected to increase, regional patterns are highly variable, with some areas becoming wetter and others drier. Extreme precipitation events are projected to become more intense.
  • Cryosphere: A significant decrease in Northern Hemisphere spring snow cover and Arctic sea ice extent is projected, particularly under high emission scenarios. Glacier melt is also expected to contribute to sea-level rise.
  • Extreme Weather: Heatwaves are projected to become more frequent, intense, and longer-lasting.
  • Carbon Cycle: Climate models indicate a positive feedback loop where a warming climate may reduce the capacity of oceans and land to absorb CO2, further accelerating warming.

Limitations and Future Work

The authors acknowledge several limitations:

  • Persistent Uncertainties: Significant uncertainties remain in climate projections, particularly concerning regional climate changes, cloud feedbacks, and the precise response of the carbon cycle.
  • Model Resolution: The limited resolution of some climate models restricts their ability to accurately simulate fine-scale phenomena like tropical cyclones.
  • Decadal Prediction Challenges: While decadal prediction is improving, understanding and predicting internal climate variability, especially in different ocean basins, remains a challenge.
  • Geoengineering: The paper briefly touches upon geoengineering techniques but notes that their feasibility, effectiveness, and potential risks are still subjects of ongoing research and debate.

Future work suggested by the paper includes:

  • Improving Cloud Feedback Understanding: This is identified as a primary source of uncertainty in climate sensitivity estimates.
  • Refining Regional Climate Projections: Further research is needed to reduce uncertainties in regional climate change predictions, especially for precipitation.
  • Integrating Socio-economic and Climate Models: Continued development of coupled models is essential for a more comprehensive understanding of climate system responses.
  • Exploring the Role of Internal Variability: Better understanding and prediction of natural climate fluctuations are crucial for both decadal predictions and long-term projections.

Summary updated

Abstract

Les modèles de climat offrent la possibilité d'anticiper la façon dont le système climatique risque d'être modifié par l'activité humaine au cours du XXIe siècle. Les études s'appuient sur des simulations numériques qui explorent l'évolution du climat moyen et de sa variabilité en fonction de différents scénarios socio-économiques. Nous présentons une sélection de résultats du projet CMIP5 d'intercomparaison demodèles de climat avec un focus illustratif sur les deux modèles français ayant participé à cet exercice. On y décrit les effets attendus des perturbations humaines sur la température de surface, les précipitations, la cryosphère mais aussi les extrêmes météorologiques et le cycle du carbone. Les résultats font ressortir de nombreux aspects fiables, que ce soit sur l'amplitude et la répartition géographique des changements attendus ou sur les processus mis en jeu dans ces changements. Ils montrent aussi les limites de l'exercice prospectif et des incertitudes persistantes sur certains points clés.
BibTeX
@inproceedings{boucher2015projection,
  author = {Oliviér Boucher and Jean‐Louis Dufresne and Jessica Vial and Éric Brun and Julien Cattiaux and Fabrice Chauvin and David Salas y Mélia and Aurore Voldoire and Laurent Bopp and Pascale Braconnot and Philippe Ciais and Pascal Yiou and Éric Guilyardi and Emmanuel Mignot and Céline Guivarch},
  title = {{Projection des changements climatiques futurs}},
  booktitle = {La},
  year = {2015},
  doi = {10.4267/2042/56362},
}
Figure