Stream 1B: Risk Management

Moderator: Simon Burchett, WSP

Question – Lluis Via-Estrem, Moody’s RMS: Regarding the model you are developing, you said that the resolution is increasing every day. What’s the current status of the model, and what is the computational cost? You mentioned a few million elements – so a model this big surely has a cost to run…? Also, you run TOMAWAC for waves, and tides. What about search and storm search? Is that a component that you include?

Answer – Giovanni Cuomo: The model I presented today is based on the TELEMAC modelling suite and is therefore extremely efficient. More in the specific, we are using TOMAWAC to simulate short waves and swells; and TELEMAC to predict circulation (and associated levels) for long waves (tides, storm surges and tsunamis). I’m always up for using the simplest model that does the job, but nothing simpler! Sometimes representing effectively the relevant physics requires going a step further and increase level of complexity to achieve a more realistic representation. This might for example require using a 3D model or increase model resolution or coupling different models (particularly in those area in which sensitivity analysis suggest that not doing so would lead to significant errors, shallow foreshores, flood plains and so forth). As for the model resolution… I wouldn’t say it is increasing day by day, but rapidly, yes. Some of the regional models presented today have a sub-kilometre resolution (i.e. 500m). We run the model on our own HPC system.