The European Centre for Medium-Range Weather Forecasts (ECMWF) has launched a revolutionary AI-powered weather model that significantly improves both accuracy and speed over traditional forecasting methods. Trained on extensive climate data, the model uses far less computing power while delivering more reliable predictions for temperatures, rainfall, and wind speeds.

The breakthrough particularly benefits energy trading firms increasingly dependent on wind and solar power forecasting. Unlike traditional ensemble forecasts using 50 weather simulations, the AI model provides faster calculations that traders can integrate into their own algorithms.

Crucially, ECMWF is releasing the model's "weights"—the AI's learned patterns—as open-source technology with commercial licensing. This allows users to bypass expensive GPU calculations and directly monetize the forecasting data.

Industry experts call it a "turning point" in operational forecasting that could spark a boom in customized weather tools for public forecasting and energy trading. Users can leverage the pre-trained atmospheric knowledge without needing to develop their own models from scratch.

A coalition of 20 states filed a federal lawsuit Wednesday challenging the Trump administration's decision to shut down FEMA's Building Resilient Infrastructure and Communities (BRIC) program. The suit, filed in Boston, argues that FEMA unlawfully terminated the multibillion-dollar disaster preparedness grant program without Congressional approval.

BRIC, established in 2018 during Trump's first term, helps communities protect infrastructure from floods, hurricanes, and other natural disasters through projects like raising roads and upgrading storm systems. Over four years, FEMA allocated $4.5 billion across nearly 2,000 projects, with the program and similar initiatives saving taxpayers over $150 billion in avoided rebuilding costs over two decades.

The states—including New Jersey, New York, California, and 17 others—argue the program's elimination puts lives at risk, particularly as communities face increasing extreme weather events. FEMA announced the shutdown in April, claiming BRIC was "wasteful and ineffective" and focused on "political agendas," though provided no supporting evidence.

Researchers from the Barcelona Supercomputing Center and European Central Bank analyzed 16 extreme weather events between 2022-2024, linking them to dramatic food price increases driven by climate change. Key examples include a 300% spike in Australian lettuce prices after record flooding, 80% higher US vegetable costs due to California's worst drought and Hurricane Ian, and 40% increases in Chinese vegetable prices after a heat wave saw temperatures reach 115°F. The study found these unprecedented weather conditions are becoming increasingly common and expensive in the near-term. British households paid an extra £361 ($484) in 2022-2023 due to climate-related food inflation. While price shocks are typically short-term as production adjusts, specialty products like coffee and cattle face longer-term impacts due to specific climate requirements.

China's Premier Li Qiang announced construction of the world's largest hydropower dam on the Tibetan Plateau, estimated to cost at least $170 billion. The project, dubbed a "project of the century," will consist of five cascade stations generating 300 billion kilowatt-hours annually—equivalent to Britain's total electricity consumption.

Located on the Yarlung Zangbo River's lower reaches, the dam capitalizes on a 2,000-meter drop over 50 kilometers. Operations are expected in the 2030s, with China viewing it as a crucial economic stimulus.

However, the project faces significant opposition. India and Bangladesh worry about downstream impacts on millions of people, as the river becomes the Brahmaputra flowing through both countries. NGOs warn of irreversible environmental damage to the biodiverse Tibetan Plateau. India's Arunachal Pradesh chief minister fears the dam could dry 80% of the river in his state while potentially flooding downstream areas.