News and Blog

EPSRC CDT in Next Generation Computational Modelling

David Lambkin, ABPmer - Numerical Simulations of Coastal and Shelf Seas

David Lambkin gives the first seminar of the year

On 11 October David Lambkin of ABPmer gave the first NGCM seminar of the academic year.

ABP Marine Environmental Research work on coastal defences and management, energy studies for environmental impact, planning and licensing, and the development of ports and navigation. They study and predict the physical properties of the marine environment, which include the wind, waves, water levels, currents, water quality, sediment flow, and coastal evolution. Whilst it's essential to get the best quality, and largest quantity, of measured experimental data, that has limited spatial and particularly temporal coverage: for example, climate effects need decades of data, but current data typically only covers 15 years.

David outlined two crucial uses of numerical modelling for ABPmer. The first is to supplement otherwise sparse data. An example is ABPmer's work on a hindcast model and database (available at http://www.seastates.net) which provides detailed and summary hourly wave data for over 31 years to the past. This is crucial for looking at historical norms and considering the likelihood of extreme events.

The second crucial use of numerical modelling is to investigate the effects of a proposed development. The example used was a propsed offshore wind farm development near the Isle of Wight, which needed an environmental assessment. Here it was essential to quantify the key effects (for example, for the impact on the wave energy with and without the development) where there is a complex coastline and with very limited experimental data (a couple of spatial locations over a period of months, whilst the impact assessment needed detailed coastal coverage over a period of decades).

After listing the range of models, software packages and approaches that ABPmer currently employ, David outlined the challenges that his industry faces in the future. The key challenges were

  • Data. Getting good data input for bathymetry and wind is essential but hard, and expensive;
  • Sofware. Crucial to their modelling, but requires continuous updating;
  • Hardware. Speed is key, but they tend to use off-the-shelf components. Currently using a lot of GPU computing, their main challenge is how the data volumes (databases of several terabytes) impact on version control and backup.
  • Analysis. As their money comes from presenting their results to the marine industry, they need to continuously improve how their data is presented.

Hans Fangohr and David Lambkin explaining the importance of numerical modelling

Finally, David talked about the ongoing development projects. He highlighted the work that Rory Brown, a cohort 1 NGCM student, did in his summer project. Taking the SPH techniques he will work on in an astrophysical context for his PhD, he studied their impact and use in coastal defence problems with ABPmer. There are ongoing opportunities to apply academic numerical modelling techniques in other marine contexts.

Comments