Improving the forecasting of storms
Working with the Met Office, our researchers have improved prediction and planning for extreme local weather.
Particular benefit [of the research] is found in predicting low-level temperatures, which is very important for maintaining the road network in a safe condition and for predicting fog.
Problems in local forecasting
Storms, fog and ice can have significant social impact – fatalities, damage to property, and paralysis of transport networks.
So accurate forecasting is essential for local organisations to take action effectively.
Forecasting local storms and fog requires high resolution data for environmental variables such as wind speed, atmospheric pressure, air temperature and moisture content.
But the computational grids (meshes) used by the UK’s Met Office to compile weather forecasts can lack sufficient resolution for local forecasting.
New cutting-edge approach
The University of Bath’s Department of Mathematical Sciences is at the forefront of research in adaptivity.
This is a sophisticated mathematical approach to automatically reducing computational mesh sizes at specific points of interest, increasing resolution while minimising any escalation in computing power.
By working closely with the Met Office, our research has led to improved procedures for numerical weather prediction and data assimilation.
The Met Office has incorporated adaptive meshes into its Operational Code for Data Assimilation, and when this is used to predict ground surface temperatures it has had an immediate impact on analyses used to advise councils on road gritting requirements.
Immediate benefits to local councils
The project benefited the Met Office’s numerical weather prediction and data assimilation.
This had an immediate impact on analyses to better advise County Councils on when to grit roads.
The project has strengthened the University’s relationships with the Met Office, leading to a further, highly successful project with the Met Office to develop a network in climate modelling.
This research was part of our REF 2014 submission for Mathematical Sciences.