Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Pastimes : Hurricane and Severe Weather Tracking -- Ignore unavailable to you. Want to Upgrade?

To: Wharf Rat who wrote (25904)	10/23/2025 3:53:04 PM
From: miraje	Read Replies (2) \| Respond to of 25958

Looks like you're in the bull's eye for the next atmospheric river to hit the west coast over the weekend.. foxweather.com Atmospheric river forecast to bring deluge of rain, snow and potential flooding to Pacific Northwest ...According to the FOX Forecast Center, new forecasts suggest that the more intense portion of the storm may have shifted farther south, with the strongest impacts targeting southern Oregon and Northern California. Cities including San Francisco and Eureka in California and Medford, Oregon, will see the most intense part of the storm on Saturday and Sunday...

To: Wharf Rat who wrote (25904)	10/23/2025 6:39:18 PM
From: Don Green	Respond to of 25958

Thoughts for Grok Ai AI in Weather Prediction: An Overview Artificial intelligence has transformed weather prediction in recent years, offering faster, more efficient alternatives to traditional physics-based models. Traditional numerical weather prediction (NWP) relies on supercomputers to simulate atmospheric physics, which can take hours or days for high-resolution forecasts. AI models, particularly those using machine learning techniques like neural networks, learn patterns from historical data to generate predictions in minutes or seconds, often with comparable or superior accuracy for certain metrics. Key advancements include: GraphCast by Google DeepMind: Released in 2023, this model uses graph neural networks to predict global weather up to 10 days ahead. It outperforms traditional models like ECMWF’s HRES in 90% of tested variables, running on standard hardware rather than supercomputers. Pangu-Weather by Huawei: A 2023 model that employs 3D Earth-specific transformers for high-resolution forecasts. It’s particularly strong in predicting extreme events like typhoons, achieving better results than conventional methods in some cases. FourCastNet by NVIDIA: Focuses on ensemble forecasting, generating multiple possible outcomes quickly to account for uncertainty. It’s used for short-term predictions and has been integrated into operational systems. AIFS (Artificial Intelligence/Integrated Forecasting System) by ECMWF: Launched in 2024, this combines AI with traditional models for medium-range forecasts, improving efficiency and reducing energy consumption. These AI systems excel in medium-range forecasting (3-10 days) and are being tested for nowcasting (immediate short-term predictions) and long-term climate projections. Benefits include lower computational costs, better handling of big data from satellites and sensors, and potential for real-time updates. Challenges remain, such as ensuring “trustworthiness”—avoiding hallucinations (false predictions), incorporating physical constraints, and handling rare events where training data is sparse. Organizations like the World Meteorological Organization (WMO) and NOAA are increasingly adopting AI, with pilot programs showing up to 20-30% improvements in forecast accuracy for precipitation and temperature. The Claim About AI2ES Termination Yes, the NSF AI Institute for Research on Trustworthy AI in Weather, Climate, and Coastal Oceanography (AI2ES)—a collaborative effort across seven universities led by Amy McGovern at the University of Oklahoma—was effectively terminated when its funding was not renewed in June 2025. 6 Launched in 2020 with a $20 million NSF grant over five years, AI2ES focused on developing reliable AI for environmental sciences, including weather prediction applications like forecasting cold-stunning events for sea turtles (saving costs by optimizing bay closures), hail storms, hurricanes, and visibility improvements for aviation and vehicles. 6 The institute emphasized “trustworthy” AI, addressing biases (e.g., urban-centric storm reporting) and integrating risk communication for better societal impact. The Office of Management and Budget (OMB) stated that the program was cut because it “wasted taxpayer funds to place climate change hysteria in AI models.” 0 This occurred amid the Trump administration’s broader push for AI investment, but also proposed 55% cuts to the NSF’s budget for FY26, which funds diverse scientific research from STEM education to biotechnology. 6 Current and former NSF staff, including Kevin Johnson, suggested the decision was likely anticipatory of these budget reductions rather than the OMB’s stated reason, as threats of cuts influenced agency-wide spending restraint. 6 Regarding “Alan’s take” that “it strikes me as completely insane,” this appears to reference Alan Gerard, former director at NOAA’s National Severe Storms Laboratory (not affiliated with AI2ES). He commented that such research is complex and time-intensive to develop, making it difficult for others to quickly replicate efforts for lifesaving decisions on events like tornadoes and floods. 6 However, the exact phrasing “completely insane” doesn’t appear in available sources; it may be a paraphrase or from an unverified commentary. The termination could disrupt AI talent pipelines, as AI2ES trained diverse students and advanced practical tools, potentially slowing U.S. progress in AI-driven weather resilience amid rising extreme events.