US weather scientists enlist support from Maths experts

Mathematicians from the University of Reading are joining forces with experts from across the USA to improve the accuracy of weather forecasts.

Announced by the Biden-Harris administration, the $6.6 million project funded by NOAA (the National Oceanic and Atmospheric Administration) will establish the Consortium for Advanced Data Assimilation Research and Education (CADRE). The group will use enhanced numerical weather prediction tools to improve forecasts.

Led by the University of Oklahoma, the project will involve the work of the Data Assimilation Research Centre (DARC) at the University of Reading. Data assimilation is the science that combines weather observations with numerical models. It is used every day to provide the starting points for weather forecasts.

Professor Sarah Dance, of the University of Reading’s School of Mathematical, Physical and Computational Sciences and co-director of DARC, said: “We are delighted that NOAA has funded a data assimilation research consortium that will carry out exchanges with the University of Reading, and are looking forward to new collaborations with our US colleagues. Weather forecasting today is better than ever before, but with new machine learning techniques we can make the computer models that predict weather even more accurate. Better weather forecasting will help us predict hazardous events including, storms, high winds, floods, droughts and heatwaves more accurately. With climate change, these events are expected to become more frequent and severe, so better predictions are critical to help protect lives and livelihoods.”

Professor Amos Lawless, also of the University of Reading’s School of Mathematical, Physical and Computational Sciences and co-director of DARC, said: “The UK spends more than £50 million on satellites and their data each year for weather forecasting. Research in data assimilation is fundamental if we are to make the best use of new observations from satellites and other instruments to improve weather forecasts. By combining weather observations with models, we can create better starting points for forecasts and solve physical and mathematical equations to predict how weather changes over time. However, there are big challenges to improving this process due to the complexity of the models and the large amount of data involved.“

Across the pond

PhD students and postdoctoral researchers from the USA will work with colleagues from the University of Reading on developing new data assimilation techniques as part of the programme. The University of Reading is also working with other UK universities, the National Centre for Earth Observation and the Met Office to design an aligned Transatlantic Data Science Academy, a skills and career-development programme in weather and climate data science that will diversify and strengthen the workforce.   

Next-generation data assimilation faces significant challenges as there are serious gaps in the global data assimilation workforce and a lack of sustained innovative data assimilation research.

CADRE will address these challenges through close collaboration with the United States’s National Oceanic and Atmospheric Administration, the UK Met Office and its university partners, including the University of Reading, and the new Transatlantic Data Science Academy.