In response to the challenges to large face-to-face events presented by COVID 19, earlier this year the EGU organisation announced their plan to create the online EGU2020 event. This was indeed a challenge given the typical 16,000-strong participation at the annual event in Vienna. With a key objective to share science with their audience and beyond, this was a worthy effort. While there are, as always, many lessons to learn from new experiences, the overall feeling from JBA participants was certainly very positive. Kudos to the EGU team for taking the leap and pushing forward into new territory. In this blog we gather some JBA highlights from the event with views and insights on some of the science under discussion.
I attended the sessions ITS4.3/AS5.2,‘Machine learning for Earth system modelling’ and ITS4.1/NP4.2, ‘Big Data in Machine Learning’ as well as presenting in the HS3.4, ‘ Deep Learning in Hydrology Science’ session. It was really exciting to see the industry come together in such large numbers and with such enthusiasm even with the different format. I particularly enjoyed understanding how others in my field are using deep learning to tackle the issues with which we are faced and found it interesting to see the trends in research. Sessions were intense, like a speed exam but it was lovely answering so many questions from both academia and the industry. I am still working my way through the displays I wanted to look at more closely – there is just so much to learn.
I attended session NH9.22, ‘Climate services for insurance and adaptation’, in which Dr Valentina Noacco of Bristol University presented her work on understanding uncertainty in catastrophe models. Using JBA Risk Management’s Global Flood Model, Valentina uses Global Sensitivity Analysis to analyse which model inputs impact the uncertainty in losses. Through collaboration with JBA, Valentina has been able to open the “black box” of catastrophe models and really explore the impact of model assumptions on outputs. It was great to see the interest this presentation sparked in others who were clearly keen to understand how these techniques might be applied to other catastrophe models.
I enjoyed my experience presenting material via text chat on integrating climate data into re/insurance tools in the online session ITS5.6/NH9.22 ‘Climate Services for Insurance and Adaptation’. Attendees had the opportunity to review display material beforehand, which meant I got several thought-provoking queries – I just wish I could type quickly enough to have answered them all then and there! We discussed everything from climate change trends in observations to uncertainty in future risk projections. With over 100 participants from industry and academia, it proved to be a very dynamic session!
Having not had the chance to attend the EGU in person in Vienna before, the prospect of it being online and accessible to everyone was very appealing. I did not think twice before getting involved. Throughout the week I joined online chats for a variety of sessions. One interesting session, NH1.1, focused on the challenges and uncertainties of flood risk modelling and assessment. As a Flood Mapping specialist at JBA, hearing other views across the science community particularly on the topic of 'uncertainty' in flood risk modelling was really insightful.
The preparation for this year’s EGU was a bit different compared to traditional conferences. As we did not get the chance to present in person, we included more text in the display and tried to clearly explain the concepts and methods. The chat session was very fast-paced and unfortunately, I did not have enough time to answer all the questions. However, the chat format made it possible to exchange ideas as compared to simply posting comments to the displays. The convener for my session made a very good job of coordinating the online conversation.
Emma Raven, Head of Research and Development at JBA Risk Management, coordinated the JBA participation. She commented,
I was never going to be able to travel to Vienna this year due to family commitments, so I really benefited from the EGUOnline experience. It was great. The online chat format packed some punch in getting a huge amount of insight across in short period of time. The material remains available and I continue to work my way through some of the interesting displays. As Head of R&D, I’m responsible for promoting academic conferences to my colleagues. Sharing science, getting feedback and learning from others is really important to our development, but we are always mindful to consider the audience at conferences – do others from our industry attend the EGU? A quick scan of the affiliations of the 100+ that attended one online chat session suggests that at least 25% of that audience work directly in the industry. It was excellent to see such a high industry presence – perhaps they are normally there just disguised in their jeans!
As Emma confirmed, the EGU sharing is not yet over. All displays are still available for review online until May 31. You can look at the comments and responses and post comments or questions yourself. Enjoy the connection with fellow scientists!
Well done EGU team and all participants.
Links to JBA Risk Managment displays at EGU 2020:
Catharine Brown, Scientific Software Developer – Author: Using image-based deep learning to identify river defences from elevation data for large-scale flood modelling
Sarah Jones, Climate Risk Consultant – Author: Assessing the global risk of climate change to re/insurers using catastrophe models and hazard maps
Valeriya Filipova, Hydrologist – Author: Regional flood frequency estimation for the contiguous USA using Artificial Neural Networks (ANN)