satellite lrHigh winds and heavy rain are familiar features of the weather in north-west Europe. Forecasting these events accurately is of the highest priority for the national weather services, and indeed forecasts have improved considerably in recent years. But there is more to do – in particular, larger-scale storm systems that may be well forecast typically contain localised regions of particularly severe weather. These smaller regions, where much of the storm damage is often concentrated, are much more difficult to forecast than the storm itself, especially more than a day ahead.

This was the challenge facing DIAMET (DIAbatic influences on Mesoscale structures in ExTratropical storms), a project involving University groups from Manchester, Leeds, Reading and East Anglia, together with the Met Office as well as NCAS and the National Centre for Earth Observation. DIAMET was part of the Natural Environment Research Council's Storm Risk Mitigation research programme, and was led by NCAS. It was a three-year project beginning October 2010 with a number of specific objectives aimed at improving weather forecasts.

At the heart of DIAMET was the way in which local heating and cooling affects the dynamics of storms. For example, as warm moist air ascends in a storm, it forms clouds of water droplets and then ice particles. Rain (and snow) forms through the growth and aggregation of these particles. But as water changes phase from vapour to water to ice, the heat energy that went into evaporating the vapour from the ocean surface is released – changing the temperature of the air. This sets up temperature gradients in the atmosphere that can have profound consequences for the development of storms. Capturing correctly the way the different particles behave, and the feedback between these processes and the storm itself, is a challenge that requires both state-of-the art weather forecasting models and detailed measurements.

The main tool that DIAMET used for the detailed measurements was the FAAM aircraft, which can fly into cyclonic storms, making detailed measurements of the temperature, humidity and wind distribution as well as the cloud and precipitation particles. When combined with satellite and ground-based radar measurements, DIAMET scientists gained a powerful insight into exactly what is happening inside the storms.

To interpret these measurements DIAMET used state-of-the-art numerical weather prediction models, in particular the Met Office's Unified Model. Models are used to provide detailed simulations of the observed storms, and DIAMET's aim was to improve those simulations, e.g. by improving the way the model represents detailed processes like the freezing of cloud droplets, convection that starts well above the Earth's surface, and the input of heat and moisture from the ocean; improving the way detailed observations are used to initialise a model forecast; and helping to distinguish which of many possible model forecasts are likely to be correct.

DIAMET conducted four aircraft campaigns in September and November-December 2011, and May and July-August 2012. This allowed the team to investigate a range of weather events in different seasons. DIAMET got off to very good start in September 2011, capturing several weather systems in the west and north of the UK. One of these in particular gave a prolonged period of heavy rain in the South-West and Midlands which proved particularly difficult to forecast accurately. Our good fortune continued in November-December with six more IOPs during a spell of particularly stormy weather. The highlight was our sampling of the very high winds over Scotland in Cyclone Friedhelm on December 8th.The summer of 2012 was one the wettest ever over the UK, giving DIAMET the opportunity for further flights in some of the storms that battered the UK that year.

DIAMET contributed to the international THORPEX research programme.