Data from the DIAMET (DIAbatic influences on Mesoscale structures in ExTratropical storms, NE/I005234/1) project, part of the Storms Risk Mitigation NERC (Natural Environment Research Council) research programme 2009-2014. DIAMET used the FAAM (Facility for Airborne Atmospheric Measurements) BAE-146 aircraft, ground-based and airborne instruments and radar together with modelling studies to forge a better understanding and prediction of mesoscale structures in synoptic-scale storms. This is determined by focusing an incident laser beam on particles, and whenever a particle passes through the beam, a shadow is generated and imaged onto the diode array. These images are part of this dataset along with flight summaries.

The DIAMET project aimed to better the understanding and prediction of mesoscale structures in synoptic-scale storms. Such structures include fronts, rain bands, secondary cyclones, sting jets etc, and are important because much of the extreme weather we experience (e.g. strong winds, heavy rain) comes from such regions. Weather forecasting models are able to capture some of this activity correctly, but there is much still to learn. By a combination of measurements and modelling, mainly using the Met Office Unified Model (UM), the project worked to better understand how mesoscale processes in cyclones give rise to severe weather and how they can be better represented in models and better forecast. The project is organised into three sections.

Real mesoscale structures in the atmosphere have been examined, using high-resolution in situ and radar measurements to derive their morphology and dynamics. The key to the latter is to calculate the production of potential vorticity by diabatic processes - especially phase changes of water (vapour/liquid/ice) and air-sea fluxes of sensible and latent heat. The associated high-resolution modelling programme will use the UM to simulate a representative number of events, diagnosing the PV tendency in the model and comparing with the measurements. Sensitivity studies and further diagnostics with the model will reveal the sensitivity of the forecasts to the correct representation of these processes and the dynamical consequences of diabatically-generated PV, both on the mesoscale and larger scales. Two student projects have investigated the role of boundary-layer processes in storm behaviour and conduct a statistical investigation of mesoscale precipitation features, based on archived radar and wind profiler data.

Examination of particular physical processes and the way these are represented in forecast models. Convection cannot be explicitly represented in current large-scale models (it is just beginning to be resolvable by high-resolution local-area models) so it needs to be parameterised. The schemes that are used are not optimised for mid-latitude storms, where convection often initiates at altitude rather than at the Earth's surface. A combination of novel diagnostics and new (or modified) schemes aimed at improving the representation of convection will be developed. Also addressed here will be the derivation of air-sea fluxes of heat and momentum from aircraft flights, and their use (as part of a larger, ongoing international project) to derive a better parameterisation for these quantities in high wind conditions. Lastly, microphysical measurements made with the FAAM aircraft will be used to derive latent heating/cooling rates as a function of the microphysical environment and used to improve the model simulations in the first WP and to improve microphysical parameterisations in the UM.

The problem of predictability will be addressed using a combination of ensemble and data assimilation techniques. A unique archive of forecast ensembles produced at the Met Office will be exploited to determine how well the forecast ensemble actually generates realistic mesoscale features, and the skill with which this is done (using standard measures of skill). Model errors in representing convection, air-sea fluxes and microphysics will be investigated to determine their impact on the forecasts for different flow conditions. The relationship between different model variables on the mesoscale is poorly known at present and this will be investigated using ensembles and the results of the measurement programme. Finally, novel approaches to data assimilation will be investigated.