07-12, 14:00–14:10 (Europe/Amsterdam), While Loop (4.2)
Using Julia for Earth and climate science has the potential to combine the best of both worlds: The speed of Fortran and the interactivity and productivity of Python, empowering users to be developers and developers to be users. In this minisymposium speakers will present software projects both from a user and a developer perspective. Talks are encouraged to discuss both use cases of existing software as well as the development of user-friendly software.
Solving the two-language problem, Julia blurs the line between users and developers. Julia software in Earth and climate science often spans multiple aspects including models and their development, preparing and running simulations, data post-processing, analysis and visualisation. Users can easily extend packages, increasing the number of contributors, while developers enhance the user interface and experience within a unified language framework. At the same time, developers start to support other software packages and tools directly given the single language framework, further growing the ecosystem.
Traditionally, Earth and climate science relied on Fortran for performance-critical software and, more recently, on Python for data analysis and visualisation. Unfortunately, this puts a divide between developers and users, at least partially counteracting the gained productivity through Python. Julia as an emerging language framework has allowed this community to rethink how to build software with and for users, of which we want to highlight efforts from both the developer but especially the user perspective. Overall we want to accelerate research on Earth and climate science, which not just depends on high-performance computing but particularly on the user interface being accessible, intuitive, interactive and extensible.
We invite both developers and users, as well as individuals who bridge these roles, to discuss software projects and their impact on user empowerment. We also welcome talks by developer-user pairs highlighting the co-benefits from a combined perspective. For this minisymposium, we accept abstracts on every field in and adjacent to Earth and climate sciences, including but not limited to
- Atmospheric sciences, including atmospheric dynamics, physics, chemistry and climatology
- Oceanography including sea ice and ocean biogeochemistry
- Earth science including land surface processes, hydrology and glaciology
- Land vegetation and interactions of ecosystems with the climate
- Geodynamics, seismology, geodesy and geochemistry
- Human and physical geography
- Climate policy and economics including integrated assessment models
Talks can focus on software that solves equations, analyses or visualises data, processes measurements and especially software with any combination of the above. Talks on planetary sciences are equally welcome especially when they parallel the issues and opportunities to the fields mentioned.
Schedule and confirmed talks (last updated 1 July)
- 14:00 Introduction
- 14:10 Hydrological modelling in Julia with Wflow.jl and Ribasim.jl
- 14:20 Multiobjective stochastic storage design of rainwater harvesting
- 14:30 High Resolution Flux Estimation with TurbulenceFlux.jl
- 14:40 SpeedyWeather.jl: Reinventing atmospheric models
- 14:50 Oceananigans.jl: An ocean-flavoured fluid dynamics library
- 15:00 Subzero.jl Fast and Flexible Sea Ice Physics
- 15:10 Subglacial water flow simulations on GPUs
- 15:20 ODINN.jl: Multi-language Geoscientific Machine Learning
- 15:30 (10min break)
- 15:40 Strategies to Integrate Data and Biogeochemical models
- 15:50 SpeciesDistributionModels.jl: an SDM workflow
- 16:00 How Climate Economics Domain Experts Use an Embedded DSL
- 16:10 Visualization and analysis of 3D Lightning Mapping Array data
- 16:20 Supporting development and operation of Scatterometers
- 16:30 Julia Community Reshapes Computation in Geosciences
- 16:40 Concluding discussion until 17:00.
Chairs
- Skylar Gering (Caltech)
- Francesco Martinuzzi (Leipzig)
- Jordi Bolibar (Grenoble)
- Milan Klöwer (MIT)
Milan is a postdoctoral associate in climate modelling at the Massachusetts Institute of Technology. He received his PhD from Oxford working on low-precision climate computing and data compression. During his PhD, Milan established the concept of the bitwise real information content for data compression. He worked with posit numbers and stochastic rounding and invented a logarithmic fixed-point number format. He ran the first 16-bit weather and climate simulation on Fujitsu's A64FX, the CPU that powers Fugaku. He writes and maintains many Julia packages. Most recently, he wrote SpeedyWeather.jl, an atmospheric general circulation model with a focus on interactivity and extensibility to further accelerate research into computationally efficient weather and climate models.
I am currently pursuing a PhD in Physics and Earth Sciences at Leipzig University, Germany, and Valencia, Spain, as a member of the ELLIS PhD program. My research focuses on the application of machine learning in Earth systems. I am part of the team at the Remote Sensing Center for Earth System Research (RSC4Earth), working under the supervision of Prof. Miguel D. Mahecha and Dr. Karin Mora. In addition, I have an affiliation with the Center for Scalable Data Analytics and Artificial Intelligence (ScaDS.AI).
I am a research software engineer working on both sea ice modeling and computational geometry in Julia.
Postdoc researcher at IGE, Université Grenoble Alpes (France)