Workshop: Building reproducible workflows for earth sciences

Monday, 14 October
09:15 10:00
Registration
45m
10:00 10:15
Welcome and introduction
Speakers: Claudia Vitolo (ECMWF), Florian Pappenberger (ECMWF)

ECMWF

Reading
15m
10:15 10:30
Reproducible workflows - Setting the scene
Speaker: Stephan Siemen (ECMWF)

ECMWF

Reading
15m
10:30 11:00
Coffee break

ECMWF

Reading
30m
Keynote
11:00 11:30
Responding to reproducibility challenges from physics to social sciences
Speaker: Ana Trisovic (IQSS, Harvard University )

ECMWF

Reading
30m
Version control for code and data
11:30 11:45
Scientific reproducible workflows - what, when and how to automate things
Speaker: Tania Allard (Microsoft)

ECMWF

Reading
15m
11:45 12:00
TBD

ECMWF

Reading
15m
12:00 12:15
Branching models - why different approaches?
Speakers: Andrew Bennett (ECMWF), Sebastien Villaume (ECMWF)

ECMWF

Reading
15m
12:15 12:30
Versioning and tracking changes of vector data
Speaker: Saber Razmjooei (Lutra Consulting)

ECMWF

Reading
15m
12:30 13:00
Open discussion: The importance of versioning
Chair: Stephan Siemen (ECMWF)
30m
13:00 14:00
Lunch break

ECMWF

Reading
1h
The role of standards
14:00 14:15
Leveraging OGC standards to boost reproducibilty
Speaker: TBC

ECMWF

Reading
15m
14:15 14:30
Workflows with the new OGC API processes specification
Speaker: Francesco Bartoli (Geobeyond Srl)

ECMWF

Reading
15m
14:30 14:45
Challenges and needs of reproducible workflows of Open Big Weather and Climate data
Speaker: Julia Wagemann (ECMWF)

ECMWF

Reading
15m
14:45 15:00
Stanardised data representation - power of reproducible workflow
Speaker: Marijana Crepulja (ECMWF)

ECMWF

Reading
15m
15:00 15:15
ECMWF data governance
Speaker: Sebastien Villaume (ECMWF)

ECMWF

Reading
15m
15:15 15:30
Reproducible tools for streamining visualisation tasks
Speaker: Milana Vuckovic (ECMWF)

ECMWF

Reading
15m
15:30 16:00
Coffee break

ECMWF

Reading
30m
Reproducibility in production
16:00 16:15
ECMWF's new product generation - Lessons learned from development to operations
Speaker: Tiago Quintino (ECMWF)

ECMWF

Reading
15m
16:15 16:30
Developing a unified workflow for Convection-Allowing Applications of the FV3
Speaker: Benjamin Blake (NOAA/NWS/NCEP/EMC)

ECMWF

Reading
15m
16:30 16:45
Refactoring EFAS product generation - Lessons learned on testing, performance and reproducibility
Speaker: Corentin Carton de Wiart (ECMWF)

ECMWF

Reading
15m
16:45 17:00
Workflow in CESM2
Speaker: Jim Edwards (National Center for Atmospheric Research USA)

ECMWF

Reading
15m
17:30 18:30
Drinks reception

ECMWF

Reading
1h
Tuesday, 15 October
09:15 09:30
Recap from day 1 and remarks
Speaker: Claudia Vitolo (ECMWF)

ECMWF

Reading
15m
Keynote
09:30 10:00
Scaling Reproducible Research with Project Jupyter
Speaker: Carol Willing (Project Jupyter)

ECMWF

Reading
30m
Tools and environments to facilitate reprocucibility
10:00 10:15
Automated production of high value air quality forecasts with Pangeo, Papermill and Krontab
Speaker: Peter Killick (Met Office Informatics Lab)

ECMWF

Reading
15m
10:15 10:30
CROW: Python-based Configuration Toolbox for Operational and Development Workflows
Speaker: Kate Friedman (NOAA)

ECMWF

Reading
15m
10:30 11:00
Coffee break

ECMWF

Reading
30m
11:00 11:15
Keeping track of data (geo)science experiments with DVC
Speaker: Kevin Donkers (Met Office)

ECMWF

Reading
15m
11:15 11:30
CMIP6 post-processing workflow at the Met Office
Speaker: Stephen Haddad (Met Office)

ECMWF

Reading
15m
11:30 11:45
Earth System models and workflows reproducibility in the BSC
Speaker: Miguel Castrillo (BSC-CNS)

ECMWF

Reading
15m
11:45 12:00
Space situational awareness - virtual search environment
Speaker: Marek Kubel-Grabau (Eversis)

ECMWF

Reading
15m
12:00 12:15
Building robust and reproducible workflows with Cylc and Rose
Speaker: Stuart Whitehouse (Met Office)

ECMWF

Reading
15m
12:30 13:00
Open discussion: Best practice for collaborative and reusable work
30m
13:00 14:00
Lunch break

ECMWF

Reading
1h
Reproducible machine learning for Earth System applications
14:00 14:15
Machine learning for flood applications
Speakers: Lukas Kugler (University of Vienna), Sebastian Lehner (University of Vienna)

ECMWF

Reading
15m
14:15 14:30
Machine learning for drought applications
Speakers: Gabriel Tseng (Okra Solar), Thomas Lees (Oxford University)

ECMWF

Reading
15m
14:30 15:00
Scaling Machine Learning with the help of Cloud Computing
Speaker: Salman Iqbal (ONS / Learnk8s)

ECMWF

Reading
30m
15:00 15:15
TBD

ECMWF

Reading
15m
15:15 15:30
TBD

ECMWF

Reading
15m
15:30 16:00
Coffee break

ECMWF

Reading
30m
Gaining flexibility through containers
16:00 16:30
Introduction to containers + demos
Speaker: Carlos Valiente (ECMWF)

ECMWF

Reading
30m
16:30 16:45
The Role of Containers in Reproducible Workflows
Speaker: Shane Canon (Lawrence Berkeley National Lab)

ECMWF

Reading
15m
16:45 17:00
Singularity containers
Speaker: Gregory Kurtzer (SYLAB)

ECMWF

Reading
15m
19:00 21:00
Workshop dinner
2h
Wednesday, 16 October
09:15 09:30
Recap from day 2 and remarks
Speaker: Claudia Vitolo (ECMWF)

ECMWF

Reading
15m
Keynote
09:30 10:00
Publishing Reproducible Geoscientific Papers: Status quo, benefits, and opportunities
Speaker: Markus Konkol (University of Münster, Institute for Geoinformatics)

ECMWF

Reading
30m
Platforms for scalable and reproducible science
10:00 10:15
DARE: Integrating solutions for data-intensive and reproducible science
Speaker: Alessandro Spinuso (KNMI)

ECMWF

Reading
15m
10:15 10:30
TBD
Speaker: Ana Trisovic (IQSS, Harvard University )

ECMWF

Reading
15m
10:30 11:00
Coffee break

ECMWF

Reading
30m
The challenges of reproducibility for cloud-based application
11:00 11:15
Current cloud activities and projects of the Centre
Speaker: Vasileios Baousis (ECMWF)

ECMWF

Reading
15m
11:15 11:30
A journey into the long white Cloud
Speaker: Andy Ziegler (Meteorological Service NZ Ltd.)

ECMWF

Reading
15m
11:30 11:45
Using Cloud to Streamline R&D Workflow
Speaker: Roope Tervo (Finnish Meteorological Institute)

ECMWF

Reading
15m
11:45 12:00
CDS and toolbox
Speaker: Cedric Bergeron (ECMWF)

ECMWF

Reading
15m
12:00 12:15
Using Jupyter notebook to interact with CDS

ECMWF

Reading
15m
12:15 12:30
TBD

ECMWF

Reading
15m
12:30 13:00
Open discussion: Reproducible parallel computations
30m
13:00 13:30
Closing
30m