Using History Matching to Improve the Accuracy of Seasonal Forecasts

Meteorological experts at the world-leading Institute for Environmental Analytics (IEA) in the UK are using novel data science techniques to help fresh produce companies improve crop management and sourcing via a new cloud-based data platform – WeatherAsset.

According to Colin McKinnon, CEO, traditional weather forecasts are often unreliable when looking further out than a week. To enable growers and retailers to see further into the season, the IEA analysts combine the best short-term weather forecasts with a computational history-matching technique to draw lessons from similar seasons from the decades-long archive of data sitting behind WeatherAsset. These are then used to show what is most likely to happen over the next six weeks. 

The new seasonal forecasts help customers answer questions like “Would we expect rainfall to increase this late in the season?”, “Are warmth units likely to reach their target for a harvest by the end of the month?” and “What are the likely yields by harvest time?”

The new platform is designed to save retailers, growers and importers time and effort in assessing how the weather might impact crop prices, quality and supply. The team believe that providing high quality global weather data and insight in an easy to use format will help customers to manage risk and plan ahead as more variable weather patterns disrupt supply chains in future.

You can find out more information about WeatherAsset by visiting their stand at this year’s online Fruit Attraction LIVE trade show during October or by clicking on www.weatherasset.com