Source proof for The Met Office Makes Leaps in Last Mile Forecast Generation with Amazon Nova

The Met Office Makes Leaps in Last Mile Forecast Generation with Amazon Nova | Case Study | AWS Skip to main content Filter: All English Contact us AWS Mark…

…particularly with the vision language opportunity.” For its LLM, the Met Office used Nova Pro in Amazon Bedrock with maritime-specific prompt engineering to turn gridded data into sophisticated text descriptions and forecasts. For its VLM, it fine-tuned Nova Lite using Amazon SageMaker distributed traini…

…against equivalent human-written and issued forecasts over a three month period demonstrated 62 percent LLM accuracy for complete forecast generation and 52 percent VLM accuracy in data to text conversion across a number of diffe…

…to text conversion across a number of different weather components. Meanwhile, it saw 83 percent gale warning accuracy compared with forecasters’ observations. While already impressive, the team anticipate that this performance may be fur…

…ntegrity, the organization collaborated closely with the AWS Prototyping team. “It took just four weeks from spinning up the environment to the final set of initial experiments. It is hugely impressive what was achieved in that time by leveraging our combin…

…n, and free meteorologists to focus on complex weather-related decision-making. About Met Office Founded in 1854, the Met Office is the UK’s meteorological service with a reput…

…s on complex weather-related decision-making. About Met Office Founded in 1854, the Met Office is the UK’s meteorological service with a reputation for continually pushing the boundaries of scientific, technological, and operational expertise. It provides critical weather and climate data and insights to help public poli…

…explains Steele. Solution | Automating maritime forecasting in just four weeks As a government agency, the Met Office is trusted to maintain a high bar of accuracy, meaning text generation must meet strict quality and consistency standards, an…

…ipt is manually generated. For some of the key marine warning’s services alone, expert meteorologists spend approximately 2,920 hours a year evaluating vast datasets spanning different weather variables, from the winds to waves, visibility, and other weather types across 31 different sea areas. They must then accurately condense data into a prescribed text format for broa…

…the release of Amazon Nova, particularly with the vision language opportunity.” For its LLM, the Met Office used Nova Pro in Amazon Bedrock with maritime-specific prompt engineering to turn gridded data into sophisticated text descriptions and forecasts. For its VLM, it fine-tuned Nova Lite using Amazon SageMaker distributed trainin…

…eering to turn gridded data into sophisticated text descriptions and forecasts. For its VLM, it fine-tuned Nova Lite using Amazon SageMaker distributed training. By doing so, it could process hourly weather evolution through visual pattern r…

…through visual pattern recognition. Rather than manually producing statements, the resulting prototype was a single simplified workflow that can easily be customized to other use cases. As the first example of custom fine-tuning Amazon Nova for vision capabilities…

…rformance, particularly given the exactness of the word-based evaluation used.” Strict word-level comparisons—explicitly counting the exact number of matched words (true positives), missed words (false negatives), extra words (false positives)—against equivalent human-written and issued forecasts over a three month period demonstrated 62 percent LLM accuracy for complete forecast generation and 52 pe…

…eriod demonstrated 62 percent LLM accuracy for complete forecast generation and 52 percent VLM accuracy in data to text conversion across a number of different weather components. Meanwhile, it saw 83 percent gale warning accuracy compared with forecasters’…

…rict quality and consistency standards, and the experiments conducted therefore represent an early (albeit significant) milestone in exploring how AI could be used across its numerous products and services. To innovate rapidly while maintaining integrity, the organization collaborated…

…seful data and intelligence. This project absolutely sits at the core of that.” Further refinement and testing is required before any potential AI-based improvements to the Shipping Forecast can be made operational, but the project demonstrates remarkable potential that could mean improvements…

…its training data set and refining prompts for machine interpretation. Because almost 300 of its current products and services involve transforming raw grid data to text, the Shipping Forecast discovery pathfinder provides a valuable foundation for possible efficiencies and automation across its portfolio. As Steele notes, “The infrastructure and scalable pattern for developing stand…

…the Shipping Forecast prototype considered a range of candidate approaches—from leveraging a vision language model (VLM) for meteorological data-to-text conversion, encoding gridded weather data in 24-frame video for direct vision processing,…

…ncoding gridded weather data in 24-frame video for direct vision processing, to using a large language model (LLM) benchmark, that processed data through an intermediate text representation stage prior to…

…e organization is using AI to forge a path forward in weather forecasting using Amazon Nova Foundation Models. Benefits large language model (LLM) accuracy 62% vision language model (VLM) a…

…e Aurora Serverless relational database service for PostgreSQL, MySQL, and DSQL Amazon Bedrock The end-to-end platform for building generative AI applications and agents Amaz…

…er prices Single-product pricing Amazon S3 Amazon Bedrock Amazon RDS Amazon EC2 Amazon SageMaker Browse all product pricing Learn Documentation Get detailed technical guides an…

…th forward in weather forecasting using Amazon Nova Foundation Models. Benefits large language model (LLM) accuracy 62% vision language model (VLM) accuracy 52% gale warning accuracy 83% weeks from c…

…n shaping its future in collaboration with the AWS Specialist Prototyping Team, it developed a prototype for turning raw data into readable forecasts using Amazon Nova Foundation Models . This discovery pathfinder project is paving the way for traditional weather s…