climate modellers

Spanish politics is in focus around Valladolid, where attention has turned to Prime Minister Pedro Sánchez's appeal to left-wing voters for a 'useful vote' for the PSOE. Sánchez framed his party as the guarantor of the welfare state, contrasting it with the model of the PP and Vox, which he said 'leaves people out in the cold'. The message lands amid a contested electoral climate in Spain.

New research published in Nature reports that high wind speeds reduce water use efficiency across terrestrial ecosystems, meaning plants lose more water relative to the carbon they absorb under windy conditions. The finding suggests wind is an underappreciated factor in how vegetation and ecosystems respond to climate change, with implications for models of carbon and water cycles.

Phys.org has published a Q&A exploring how climate models could be expanded to better capture the complexities of public health. The discussion highlights that current models focus largely on physical climate variables, while health outcomes depend on factors like disease transmission, heat vulnerability and infrastructure. Researchers argue that integrating health data into climate projections would help governments anticipate risks and plan responses.

Researchers have proposed a new weighting framework designed to help climate models identify their most reliable physics settings. The method reportedly improves how different model configurations are evaluated, allowing scientists to select the parameter settings that best reproduce observed climate behaviour. The announcement has drawn attention from climate science observers interested in refining model projections.

Kent Pitman has published a new blog essay, The Triage Problem, arguing that a critical modularity flaw lies at the heart of systematic overconsumption of planetary-level resources in typical large language model architectures. The piece frames the issue as a triage question: how computing resources are allocated across AI systems. It is circulating among climate-conscious technology discussants.

A University of Arizona researcher has co-authored a new study finding that ancient climate change operated differently from the warming happening today. The research suggests past climatic shifts cannot serve as a direct template for current human-driven climate change, a distinction with implications for how scientists model and communicate modern warming trends.

Google and McKinsey have agreed to a carbon removal project in Brazil based on silvopasture, a practice that integrates trees with livestock grazing. The deal would generate carbon credits by restoring vegetation on pastureland, tying corporate climate commitments to Brazilian agriculture. The announcement is drawing attention as a new model for corporate-funded carbon removal in the region.

New machine learning models that incorporate physics principles are being used to accelerate simulations of wind and atmospheric pressure, cutting the heavy computational cost of traditional numerical methods. The approach embeds physical laws directly into neural networks, allowing faster and more reliable forecasts of wind behaviour and pressure patterns. Researchers see it as a promising step toward quicker weather and climate modelling.

A new model of the Thwaites Glacier in West Antarctica, tested against 27 years of satellite data, projects the glacier could contribute roughly 1.2 inches to global sea levels by 2100 — about double earlier estimates. The findings are drawing attention because Thwaites is considered one of the most unstable glaciers in the world, and its potential collapse could drive far larger long-term sea level rise.

A University of Washington doctoral student is featured in a Q&A discussing efforts to expand climate models so they better capture the complexities of public health. The interview covers how current models often miss health-related impacts and why integrating public health data into climate projections could improve planning and response.

The World Economic Forum is asking how societies can pursue economic prosperity while staying within the planet's ecological boundaries. The question frames a core tension in current climate and economic policy debates: how to sustain growth and living standards without exceeding environmental limits. Discussion centres on sustainable development, resource use and models for a greener economy.

Attention has turned to the Montreal Protocol, the 1987 treaty that phased out ozone-depleting substances such as CFCs. Analysts say the agreement has helped the ozone layer begin recovering, with the Antarctic hole projected to heal in the coming decades. The treaty is widely cited as a model of successful global environmental cooperation, and discussions are revisiting what lessons it offers for current climate action.

Sophia Space and MIT Lincoln Laboratory have announced a research collaboration focused on space-based weather modeling. The partnership will combine the laboratory's research capabilities with Sophia Space's space infrastructure expertise to advance weather modeling from orbit. The announcement, carried by PR Newswire, signals growing commercial and academic interest in satellite-based approaches to weather forecasting and climate monitoring.

An Australian-based commentator argues that mainstream economic forecasting is built on flawed assumptions because it treats natural resources as infinite, overlooking externalities such as environmental damage. The critique says economists debate parameters and economic rationalism without accounting for the real limits of the planet's resources, a gap that is drawing attention amid ongoing debate over climate economics.

A new study in Nature finds that multiple global environmental changes, such as climate warming and land-use shifts, drive trade-offs between inorganic and organic carbon stored in soils. As conditions shift, one form of soil carbon can rise while the other declines, complicating estimates of total soil carbon storage and its role in the global carbon cycle. Researchers say the findings matter for climate models and land management, since soils are among the largest carbon reservoirs on Earth.

Verisk's Guin argues that modelling European extratropical cyclones must move beyond historical data, outlining a new scientific approach to assessing these storm risks. The piece addresses how insurers can better understand windstorm losses in Europe, where the historical record alone may not capture the full range of possible cyclone behaviour and financial exposure.

New research indicates climate models underestimate how much ice glaciers lose during heatwaves. A day-by-day glacier simulation projects 5% to 10% more ice loss by 2100 than conventional monthly models, which could add up to 1.3 inches of additional sea level rise. The findings suggest shorter model timeframes capture extreme heat events that coarser models miss.

New coverage examines the tension between artificial intelligence's growing energy use and emissions, and its possible value as a tool for climate action. The debate centres on whether AI's benefits, such as optimising energy systems and modelling climate risks, can outweigh the carbon cost of training and running large models, as data centre demand accelerates worldwide.

Controlled environment agriculture, including greenhouses and vertical farms, is being mapped across the United States to show which regions are best suited for indoor farming. The focus is on where factors such as climate, energy costs, water availability and proximity to markets make growing food indoors economically viable, as the sector seeks stable ground after recent financial setbacks.

A South China Morning Post opinion piece argues that China has made climate resilience a core component of its economic growth strategy, rather than treating it as a separate environmental concern. The article examines how policies on flood defences, renewable energy and adaptation measures have been folded into the country's broader development planning, positioning climate preparedness as an economic opportunity as well as a necessity.

Analysts at Vietnam Investment Review argue that carbon finance can only deliver climate results if it is channelled into commercially viable ventures. The piece highlights that carbon markets and credits depend on projects generating real, verifiable emission reductions with sound business models. Without bankable ventures, financing flows and the credibility of carbon trading mechanisms remain in question.

New analysis using artificial intelligence indicates that New Zealand's most extreme rainfall events will become more intense as the climate warms. The research, published in The Conversation, applies AI methods to model how heavy downpours are likely to change, offering sharper projections of future flood risk. It adds to growing evidence that extreme precipitation is worsening in a hotter climate.

A new analysis argues that mainstream climate models assume continuous economic growth, and that this assumption may be obscuring the most effective ways to cut emissions. By treating growth as a given, the models could be overlooking approaches that question consumption levels themselves, limiting the range of strategies considered for meeting climate targets.