Workshop on seasonal hurricane forecasting

BSC hosts the Workshop on seasonal hurricane forecasting on Friday 2nd of November 2018.

Date: 02/Nov/2018 Time: 09:00


Aula de Teleensenyament room:

Jordi Girona, 1-3
Edifici B3 - Campus Nord

Primary tabs

  • Identify gaps in knowledge
  • Identify opportunity for research and collaboration
  • Discuss the use of new technologies
  • Review skill of recent forecasts


    Atlantic Basin Seasonal Hurricane Prediction: Past, Present and Future

    Phil Klotzbach (Colorado State University)


    Change of venue (to Severo Ochoa room, BSC) + Coffee break


    Introductory word and objectives of the workshop (LP)


    Introduction of participants (if workshop is small enough, (<15 people) each participant could prepare 1 slide explaining where they are from and what their interest is in the topic)


    Session 1: 30 years on: What do we know and what are we missing?


    Chair: Suzana Camargo



    Key points:

    • What can we do well (if anything)?

    • Is that useful to anyone? And can we make the information more relevant/actionable?

    • What do we struggle with? Of those, which ones are

      • Most important?

      • Most achievable to answer?

      • Is there an overlap?

    • What is stopping us from making any progress?

    • How can we improve these particular aspects?

      • Targeted experiments?

      • Analysis?

    • Where do landfall forecasts fit into this picture?

    • Is there possibility/interest for collaborating on these aspects, either through a publication or through a collaborative project (e.g. H2020)?


    Lunch (Hall of MareNostrum)


    Session 2: Machine learning: application to hurricane forecasting


    Chair: Tom Philp



    Key points:

    • Introductory talk on machine learning by Nicolas Bruneau (Reask)

    • Short presentation of work done by Albert Kahira on the application of machine learning to the seasonal hurricane forecasting problem

    • Can ML be applied to solve some of the problems identified previously?

    • Can ML be used to develop a user-relevant application? If so, what would that look like?


    Coffee (Severo Ochoa room)


    Session 3: Contrasting the forecast performances over the 2017 and 2018 Atlantic season


    Key points:

    • Preliminary analysis of the 2018 Atlantic forecasts

    • Performance review of the 2017 Atlantic forecasts

    • Can we forecast extreme seasons like 2017?

    • Suggested modifications/improvements for hurricane website.


    Chair:  Phil Klotzbach



    Guided tour of the supercomputer


    Dinner (Gut)