TECHNOLOGY
JWA's CUTTING-EDGE TECHNOLOGY
FOR YOUR BUSINESS
Weather data predicts societal trends.
JWA's Weather as a Service (WaaS) revolutionizes business accuracy.
By utilizing proprietary weather models and advanced deep learning, it offers unparalleled predictive accuracy,
empowering businesses and solving critical challenges with top-tier forecasting.
Reasons why JWA leads in weather forecasting technology
Since its establishment in 1950,
Japan Weather Association (JWA) has consistently employed cutting-edge weather observation and forecasting technology.
Leveraging the forecasting techniques honed to address Japan's diverse weather conditions
such as seasonal variations with significant temperature differences and heavy rainfall
as well as natural disaster risks like typhoons and torrential rains,
JWA provides accurate and highly reliable weather information both domestically and internationally.
01
Numerical Weather Prediction*1
The Foundation of JWA’s Highly Reliable and
Precise Forecasts
Proprietary Model「SYNFOS」
JWA has developed and operates a proprietary numerical weather prediction system called "SYNFOS." This comprehensive numerical weather prediction system is based on Weather Research and Forecasting (WRF) model developed by a U.S. research institution, with unique enhancements made by JWA.
Based on data observed by the Japan Meteorological Agency, JWA predicts weather conditions using our numerical weather prediction model "SYNFOS" to produce and provide forecasts with exceptional accuracy and reliability.
*1 Numerical Weather Prediction (Numerical Forecasting): A method that applies calculation programs (numerical weather prediction models) based on physical equations to data observed in various locations, forecasting and simulating the future state of the atmosphere.
Predicting Weather Phenomena with Greater Precision
Advantages of using our unique numerical weather prediction system "SYNFOS":
1. Highly Accurate and Detailed Weather Forecasts
It provides highly precise predictions, especially for phenomena like heavy rainfall and insolation.
2. Forecasting elements not available from Japan Meteorological Agency Data: These include parameters such as wind speed at 100 meters altitude and hail.
02
「JWA Blend of Models」
A Unique Blend of Global Forecast Models
High Precision, High Frequency,
High Resolution
Developed by integrating “SYNFOS” with the world’s leading numerical weather prediction models, JWA has developed a new JWA Integrated Weather Forecast “JWA Blend of Models” to provide the most reliable and accurate predictions.
Meteorological organizations across the world, including the Japan Meteorological Agency, operate numerical weather prediction models that cover the entire world. However, each model differs in grid spacing, maximum forecast period, and specialized prediction areas.
“JWA Blend of Models” provides highly reliable weather forecasts by correcting each forecast model and integrating the models according to their forecast accuracy.
Additionally, the factors used for these corrections and integrations are updated daily, utilizing JWA’s accumulated meteorological expertise and AI to reflect latest data.
Hourly Forecasts Up to 14 Days Ahead
Key Features of the “JWA Blend of Models”:
1. Provides high-precision forecast information with a 1 km grid resolution up to two weeks ahead.
2. Frequent forecast updates: Forecasts are frequently updated with up to 24 times a day for 78-hour forecasts and 4 times a day for 345-hour forecasts.
This high-frequency updating helps users with pinpoint forecasts, proactive disaster preparedness, and various demand predictions.
Accuracy Verification
JWA continually verifies and evaluates the accuracy of weather forecasts, and shares the results on our website.
No reports available yet.
03
Introducing First Technologies in the Weather Industry
Cutting-Edge Technologies to Solve Societal and Business Challenges
High-Precision 2-Year Weather Forecast
JWA has launched a 2-year weather forecast, the longest-term weather forecast in the weather industry, with the advice and collaboration from Professor Hiroaki Ueda of the University of Tsukuba. This newly developed "2-Year Long-Term Weather Forecast" can predict weather conditions up to two years in advance and improves prediction error by 20-40% compared to conventional methods.
It also enables the prediction of extreme weather conditions, such as extreme high temperatures during cold periods*2 and record-breaking heat during hot periods*3, over a year in advance.
With this highly accurate 2-year forecast and expert consulting services, JWA supports data-driven decision-making in critical business scenarios such as long-term planning, material procurement, manufacturing, sales and promotional plannings and new product development.
*2 extreme high temperatures during cold periods: Temperatures significantly higher than usual during cold periods.
*3 record-breaking heat during hot periods: Temperatures significantly higher than usual during hot periods.
Spatiotemporal Downscaling Method
In 2019, JWA developed a technology called the "Spatiotemporal Downscaling Method" which utilizes deep learning. This innovative approach uses AI to downscale "3-hour precipitation data on a 20 km grid" to a "1-hour precipitation data on a 5 km grid," providing higher resolution data in both space and time. While methods for downscaling in the "spatial" direction using AI already existed at that time, there were no methods for downscaling in both "time and space," making this the first development of its kind in the meteorological industry by JWA.
Unlike other approaches, this method delivers rapid and a large number of predictions, and does not require high-speed, expensive computing resources like supercomputers.
Services & Solutions
Interpreting weather data leads to predicting societal trends and human activities.
JWA provides services to various companies and organizations to enhance the accuracy of their business operations at all stages.
Weather Data API
JWA’s Web API provides access to both weather forecasts and historical weather data.
By obtaining weather data via API, you can make use of it in a variety of ways, such as displaying it on apps and web services, analyzing your own data with weather data, and building forecast models.
Product Demand
Forecasting
By analyzing highly accurate weather forecast data and product data, JWA predicts future demand.
With precise product demand forecasting, businesses can take approaches that do not rely on experience or intuition to increase sales and reduce costs.