Meteorology Forecast Model Updates

When AI meets the weather prediction: improving short-term forecasts with machine learning Short-range weather forecasts (so-called nowcasts, covering 0–6 hours ahead) are critical for daily decisions – from aviation and shipping to renewable energy, agriculture, and public safety. Yet even the most advanced numerical weather prediction (NWP) models carry systematic errors that can make forecasts less reliable. A new study by Leila Hieta and Mikko Partio (2025) shows how machine learning can significantly reduce these biases – and it’s already running operationally at the Finnish Meteorological Institute (FMI). ➡️ What’s new? Instead of relying on traditional static bias-correction techniques, the FMI team used XGBoost, a gradient-boosted decision tree algorithm, to learn from past forecast–observation mismatches. The model continuously adapts to diurnal cycles, seasonal variations, and evolving error patterns in temperature, humidity, and wind. ➡️ The results are striking: Forecast errors (RMSE) dropped by 24%–29% compared to raw NWP output. The new method extends reliable bias correction up to 11 hours ahead, a major improvement over the previous 4-hour limit. Wind gust forecasts, notoriously difficult to get right, are now included in the correction process. ➡️ How it works operationally The system runs every hour, blending fresh observations with model output to generate bias-corrected nowcasts for 2m temperature, 2m humidity, 10m wind speed, and wind gusts. These corrected fields are then integrated into FMI’s Smartmet nowcast framework, which feeds public services and decision-making tools. ➡️ Why this is exciting for the field It shows how AI can add value to physics-based models, not replace them. It demonstrates that operational meteorology can integrate open-source ML tools (the code is available on GitHub) into national forecasting systems. It paves the way for future hybrid approaches where AI not only corrects biases but also drives real-time probabilistic and ensemble nowcasts. In short: this is a big step toward smarter, more adaptive, and more accurate weather forecasts. It’s also a great example of how AI can be responsibly operationalized in critical public services. 👉 Full study: Operational Machine Learning Post-Processing of Short-Range Temperature, Humidity, Wind Speed and Gust Forecasts (Hieta & Partio, 2025) https://lnkd.in/dMT9VKg6 #WeatherForecasting #AI #MachineLearning #Nowcasting #Meteorology #XGBoost #ClimateTech #FLOODAR #GOSPACELABS

Cambridge’s Aardvark AI Weather System Rivals Supercomputers—On a Desktop Revolutionizing Forecasting with Lightning Speed and Minimal Resources In a major leap for meteorological science, researchers at the University of Cambridge, in collaboration with the Alan Turing Institute, Microsoft Research, and the European Centre for Medium Range Weather Forecasts, have unveiled Aardvark Weather—an AI-powered forecasting system that delivers high-precision weather predictions using a fraction of the computing power of traditional methods. This breakthrough marks a pivotal moment in the use of artificial intelligence for global and local weather modeling. Key Highlights of the Aardvark Weather System • Fast and Efficient • Aardvark can generate weather forecasts in minutes, a process that typically takes hours on supercomputers. • It runs on a standard desktop computer, offering a thousands-fold reduction in computational demand compared to conventional models. • Advanced AI Integration • Unlike traditional forecasting, which depends on complex numerical solvers, Aardvark uses AI models to simulate atmospheric dynamics with impressive accuracy. • The system is trained to predict both global and localized conditions, bridging the gap between general and specific weather needs. • Collaborative Development and Scientific Rigor • The project received support from top-tier institutions and tech leaders, including Microsoft and the Turing Institute, adding credibility and cutting-edge expertise. • It’s part of a broader trend of AI-enhanced meteorology, alongside efforts from Google, Huawei, and others to modernize forecasting pipelines. How Aardvark Compares to Traditional Forecasting • Reduced Cost and Accessibility • Traditional systems require expensive infrastructure and expert teams, making them inaccessible to many regions. • Aardvark’s lightweight footprint makes high-quality forecasting possible in remote or resource-limited environments. • Improved Speed and Responsiveness • The system’s rapid output means forecasts can be updated more frequently, aiding in fast-developing weather events like hurricanes or flash floods. • This has the potential to save lives and minimize damage through earlier warnings and more agile disaster response. Why This Advancement Matters Aardvark Weather exemplifies the transformative power of AI in scientific computing. By delivering supercomputer-level forecasting on desktop hardware, it democratizes access to critical climate insights and sets a new standard for speed, accuracy, and efficiency in weather prediction. As climate volatility increases, such agile and scalable tools could become essential for everything from agriculture and disaster management to global logistics and energy planning. This innovation not only redefines forecasting—it redefines who can do it.

How AI is changing storm response in the U.S. — technically. Have you experienced it? Extreme weather response is no longer driven by single forecasts. It’s driven by ensembles + AI acceleration + real-time data fusion. Here’s what’s happening under the hood: AI-accelerated Numerical Weather Prediction (NWP) Deep learning models (graph neural nets, transformers) are trained on decades of reanalysis data to approximate full physics-based solvers. Result: • Inference in seconds instead of hours • Enables rapid ensemble generation (hundreds of scenarios, not dozens) This allows forecasters to update storm tracks and intensity continuously, not on fixed cycles. Multi-modal data fusion AI ingests: • Satellite imagery (GOES) • Doppler radar volumes • Ocean buoys & atmospheric soundings • Ground IoT sensors • Historical climatology Models correlate spatial-temporal patterns across modalities — something classical models struggle with at scale. Severe weather nowcasting Computer vision models detect: • Convective initiation • Tornadic signatures • Rapid intensification signals Lead times improve by 30–60 minutes for fast-forming events — which is operationally massive for emergency management. Probabilistic forecasting, not single answers ML-driven ensembles output probability distributions, not deterministic paths: • Flood depth likelihoods • Wind gust exceedance • Ice accumulation risk This feeds directly into risk-based decision systems. Infrastructure impact modeling Utilities combine AI weather outputs with: • Grid topology • Asset age & failure history • Load forecasts This enables pre-storm optimization: • Crew pre-positioning • Targeted grid isolation • Faster restoration paths Operational decision intelligence AI systems now bridge forecast → action: • When to evacuate • Where to stage responders • Which assets fail first This is no longer meteorology alone — it’s real-time systems engineering. Storms are getting more chaotic. Our response is getting more computational. AI doesn’t replace physics. It compresses it into time we can actually use. #AI #WeatherModeling #Nowcasting #ClimateTech #InfrastructureAI #DigitalTwins #ResilienceEngineering #HPC

India’s Weather Forecasting Enters a New Era! One month ago, we launched the Bharat Forecast System (BFS), India’s most advanced, fully indigenous weather prediction model. Built by the Indian Institute of Tropical Meteorology (IITM), Pune, it marks a watershed moment in India’s ability to predict and respond to extreme weather events. Over the past 10 years, under the visionary leadership of Hon’ble Prime Minister Sh Narendra Modi , India has made enormous strides in weather and climate science. What was once reliant on foreign models is now led by Indian innovation, more precise, more timely, and more accessible. The Bharat Forecast System is a reflection of this transformation. Here’s what makes it a game-changer: • Double the precision: Earlier models operated at a 12-kilometre resolution. The BFS now works at 6-kilometre resolution, enabling us to predict localised weather conditions with much greater accuracy, especially vital for hilly terrains, coastal zones, and urban areas. • Faster forecasts: BFS runs on India’s new weather supercomputer ARKA, located at IITM Pune. With 11.77 petaflops of computing power and 33 petabytes of storage, ARKA can process billions of weather data points in real time, drastically reducing the time it takes to generate a forecast. • Smarter data integration: BFS uses a blend of data from ISRO satellites, ground stations, ocean buoys, and even information from global partners. This helps the system “see” and simulate weather patterns with far greater clarity. • Open for global science: Unlike many global systems, India is making the BFS data open-access, inviting researchers from around the world to build on our model. It promotes not just national preparedness but global scientific collaboration. What does this mean for the common citizen? Better early warnings before a cyclone hits. More accurate rainfall predictions for farmers. Improved alerts for cloudbursts and flash floods in vulnerable areas. And faster, more targeted disaster response, saving time, resources, and lives. This is not just about supercomputers or models. It is about science serving society, protecting communities, and helping us plan better, from sowing crops to managing cities. With the Bharat Forecast System, India joins the front ranks of nations shaping the future of weather prediction. It is a proud milestone for our scientists and our country, and a strong step forward in building a Viksit Bharat.

In this week's column, I look at NVIDIA's new generative foundation model that it says enables simulations of Earth’s global climate with an unprecedented level of resolution. As is so often the case with powerful new technology, however, the question is what else humans will do with it. The company expects that climate researchers will build on top of its new AI-powered model to make climate predictions that focus on five-kilometer areas. Previous leading-edge global climate models typically don’t drill below 25 to 100 kilometers. Researchers using the new model may be able to predict conditions decades into the future with a new level of precision, providing information that could help efforts to mitigate climate change or its effects. A 5-kilometer resolution may help capture vertical movements of air in the lower atmosphere that can lead to certain kinds of thunderstorms, for example, and that might be missed with other models. And to the extent that high-resolution near-term forecasts are more accurate, the accuracy of longer-term climate forecasts will improve in turn, because the accuracy of such predictions compounds over time. The model, branded by Nvidia as cBottle for “Climate in a Bottle,” compresses the scale of Earth observation data 3,000 times and transforms it into ultra-high-resolution, queryable and interactive climate simulations, according to Dion Harris, senior director of high-performance computing and AI factory solutions at Nvidia. It was trained on high-resolution physical climate simulations and estimates of observed atmospheric states over the past 50 years. It will take years, of course, to know just how accurate the model’s long-term predictions turn out to be. The The Alan Turing Institute of AI and the Max Planck Institute of Meteorology, are actively exploring the new model, Nvidia said Tuesday at the ISC 2025 computing conference in Hamburg. Bjorn Stevens, director of the Planck Institute, said it “represents a transformative leap in our ability to understand, predict and adapt to the world around us.” The Earth-2 platform is in various states of deployment at weather agencies from NOAA: National Oceanic & Atmospheric Administration in the U.S. to G42, an Abu Dhabi-based holding company focused on AI, and the National Science and Technology Center for Disaster Reduction in Taiwan. Spire Global, a provider of data analytics in areas such as climate and global security, has used Earth-2 to help improve its weather forecasts by three orders of magnitude with regards to speed and cost over the last three or four years, according to Peter Platzer, co-founder and executive chairman.

Scientists at the U.S. National Science Foundation National Center for Atmospheric Research (NSF NCAR) have begun running real-time 3-kilometer (1.9-mile) experimental weather forecasts for the entire globe. This achievement, a milestone for meteorology, can lead to significant advances in worldwide weather prediction. The real-time forecasts are more fine-scale than other real-time global weather models and so detailed that they can capture individual thunderstorms around the world. Although still in the experimental phase, the research points the way toward better protecting society from extreme weather events. The NSF NCAR forecasts are so fine-scale that they consist of about 65.5 million horizontal 3-kilometer grid cells over Earth’s entire surface. At each cell, forecasts are made at 55 vertical layers ascending into the atmosphere. The latest generation of weather models is extraordinarily effective for weather prediction, and forecasts have become increasingly accurate in recent decades. But global 3-kilometer models offer the potential for improving forecast accuracy even more, especially in remote regions of the world, which is important both for oceanic flights and communities at risk of extreme rainfall and flooding. https://lnkd.in/e885dHCr #weatherforecasting #nsf #ncar #globalweatherforecasts #mpas #extremeweather #weatherprediction #windenergy #solarenergy

Weather Was the 1st System Where AI Proved Itself In 2025, weather forecasting crossed a clear threshold. Deepmind’s GenCast, an AI-based weather model, outperformed the ECMWF ensemble system, the world’s leading weather model, in formal benchmark tests. It was more accurate on 97.2% of measured variables and forecast times, up to 15 days ahead. This is the first time an AI model has beaten the best physics-based system for global weather forecasts at this range. GenCast does not use physics equations to predict the weather. Instead, it learns from decades of past weather data. Each forecast produces more than 50 possible future scenarios, showing not just what might happen, but how likely each outcome is. This matters because real-world decisions depend on understanding risk, not just a single prediction. The second break is compute and speed. Traditional weather models run on large supercomputers and can take hours to produce a full global forecast. GenCast can do the same job in minutes using AI hardware. This makes it possible to update forecasts quickly when conditions change. This does not mean traditional models are being replaced. In 2025, weather agencies are using both AI and physics-based models together. The ECMWF has introduced its AI Forecasting System (AIFS) alongside existing models. Physics helps in rare and extreme situations, while AI adds speed, pattern recognition, and better handling of uncertainty. Physics remains foundational, but it is no longer the only path to cutting edge prediction. Hybrid systems now define the frontier. Weather went first because it had decades of data and rigorous testing standards. The same method, learning system behavior and uncertainty from data applies wherever systems are too complex to simulate step by step. GenCast proved this approach works where accuracy matters and errors are costly. Other domains will follow. #AI #WeatherForecasting #GenCast #ClimateTech #ScientificComputing #SystemsModeling Further reading: ECMWF AI forecasts become operational https://lnkd.in/gPxnZdp2

Bharat Forecasting System (BFS): Redefining Weather, Resilience, and National Readiness On May 26, 2025, India launched more than just a forecasting model, it unveiled a vision for hyperlocal climate intelligence. The Bharat Forecasting System (BFS), developed by the Indian Institute of Tropical Meteorology (IITM Pune), marks a radical leap in tropical meteorology and disaster preparedness. With a 6-km grid resolution, BFS now stands as the highest-resolution global weather model, surpassing even the US, EU, and UK. This isn’t just a computational feat, it’s a policy enabler, a disaster mitigator, and a force multiplier for agriculture, aviation, and national security. What BFS is built to achieve? > 2-hour nowcasting and 10-day medium-range forecasts down to the panchayat level > 30–64% improvement in prediction accuracy > Real-time alerts for heatwaves, flash floods, hailstorms, and cyclone tracks Sectoral support: flood prevention, precision irrigation, pest management, and defense ops What Makes It Possible? At its core is Arka, a 11.77 petaflop supercomputer with 33PB of storage — crunching 6x6 km forecasts in under 4 hours. Data streams in from 40+ Doppler radars (scaling to 100), Oceansat-3, INSAT-3D, and 2,000+ ground stations. 🌍 Why it matters globally? In a warming world, BFS is not just about rainfall and wind, it’s about socioeconomic planning powered by climate foresight. It demonstrates how countries can invest in domestic capability, public science, and operational readiness, creating systems where weather data isn’t just collected but acted upon. India has placed a stake in the future of forecasting. Now the challenge and opportunity lie in scaling its use and sharing its model. #BharatForecastingSystem #ClimateIntelligence #DisasterResilience #Supercomputing #WeatherPrediction #PublicPolicy #Monsoon #Agritech #UrbanPlanning #IITM #ISRO #IndiaScience #PrecisionWeather https://lnkd.in/gD2v_79G

For those of you following developments with AI & Science, particularly around weather forecasting… At Google Research and Google DeepMind we have introduced an experimental model for tropical cyclone prediction, which can predict a cyclone’s formation, track, intensity, size and shape – generating 50 possible scenarios, up to 15 days in advance. And as we head into this year’s cyclone season, we’re partnering with the US National Hurricane center to support their forecasts and warnings. We’re publicly sharing this experimental model in Weather Lab, a new platform to access experimental weather forecast visualizations, and we hope to gather feedback and enable researchers and forecasters to leverage our models and predictions to inform their own work. You can learn more in our blog post (https://lnkd.in/geG62c2v) or this New York Times story (https://lnkd.in/gAFPbUrD).

🚨 𝐈𝐧𝐝𝐢𝐚’𝐬 𝐖𝐞𝐚𝐭𝐡𝐞𝐫 𝐅𝐨𝐫𝐞𝐜𝐚𝐬𝐭𝐢𝐧𝐠 𝐉𝐮𝐬𝐭 𝐆𝐨𝐭 𝐚 𝐌𝐚𝐬𝐬𝐢𝐯𝐞 𝐔𝐩𝐠𝐫𝐚𝐝𝐞 𝐀𝐧𝐝 𝐈𝐭'𝐬 𝐏𝐨𝐰𝐞𝐫𝐞𝐝 𝐛𝐲 𝐚𝐧 𝐈𝐧𝐝𝐢𝐚𝐧 𝐒𝐮𝐩𝐞𝐫𝐜𝐨𝐦𝐩𝐮𝐭𝐞𝐫 🚨 𝑻𝒉𝒆 𝑩𝒉𝒂𝒓𝒂𝒕 𝑭𝒐𝒓𝒆𝒄𝒂𝒔𝒕 𝑺𝒚𝒔𝒕𝒆𝒎 𝒊𝒔 𝒉𝒆𝒓𝒆 𝒂𝒏𝒅 𝒊𝒕’𝒔 𝒂𝒃𝒐𝒖𝒕 𝒕𝒐 𝒄𝒉𝒂𝒏𝒈𝒆 𝒉𝒐𝒘 𝑰𝒏𝒅𝒊𝒂 𝒔𝒆𝒆𝒔 𝒊𝒕𝒔 𝒔𝒌𝒊𝒆𝒔. 𝐈𝐦𝐚𝐠𝐢𝐧𝐞 𝐭𝐡𝐢𝐬: You're in Mumbai during monsoon. You’ve planned a crucial meeting, or maybe just a day out with family. The forecast says “partly cloudy,” but 30 minutes later, the skies open up. Umbrellas fail. Plans get ruined. We've all been there. But what if our weather updates were precise to the street, and not just the city? What if we could get an accurate 2-hour nowcast telling us exactly what the sky will look like in your area? 𝑰𝒏𝒅𝒊𝒂 𝒋𝒖𝒔𝒕 𝒎𝒂𝒅𝒆 𝒕𝒉𝒂𝒕 𝒍𝒆𝒂𝒑. Introducing the BHARAT FORECAST SYSTEM (BFS) Developed indigenously by the Indian Institute of Tropical Meteorology and launched by Earth Sciences Minister Dr. Jitendra Singh, this is India’s most advanced weather prediction system yet. 𝑯𝒆𝒓𝒆’𝒔 𝒘𝒉𝒂𝒕 𝒎𝒂𝒌𝒆𝒔 𝒊𝒕 𝒂 𝒈𝒂𝒎𝒆-𝒄𝒉𝒂𝒏𝒈𝒆𝒓: ✅ Double the Resolution: From 12 km (Global Forecast System) to a stunning 6 km grid. More detail. More accuracy. More localised insights. ✅ Powered by ARKA, India’s Beast of a Supercomputer: • 11.77 petaflops of processing power • 33 petabytes of storage Located at IITM Pune, ARKA dwarfs the older Pratyush system in speed and capacity. ✅ Real-Time Radar Integration: • Live data from 40 Doppler Weather Radars • Expanding to 100+ soon for pan-India precision • Delivers 2-hour nowcasts crucial for cities, disaster response, aviation & farmers. ✅ Made in India. For India. Built by Indian scientists, on Indian soil, to solve India’s unique weather challenges. No imports, no borrowed tech. Why it matters: Climate change is real. Extreme weather is rising. Timely and accurate forecasts aren’t just about convenience - they’re about saving lives, crops, and businesses. The BFS is more than a TECH UPGRADE. It’s India taking control of its skies with 𝒔𝒄𝒊𝒆𝒏𝒄𝒆, 𝒔𝒑𝒆𝒆𝒅, 𝒂𝒏𝒅 𝒔𝒆𝒍𝒇-𝒓𝒆𝒍𝒊𝒂𝒏𝒄𝒆. PS: Follow Lovish Anand for updates relating to Markets, Business News, Technology, Mutual Funds and Insurance. #BharatForecastSystem #ARKA #Supercomputer #WeatherTech #MadeInIndia #WeatherForecast #ClimateTech #DigitalIndia #Nowcast #AIinWeather #India #explorewithlovish #trending Indian Institute of Tropical Meteorology Linkedin News LinkedIn News India LinkedIn Guide to Creating LinkedIn Guide to Networking LinkedIn for Learning