This topic is where you can introduce yourself to the other members of the Machine Learning for Climate and Weather Working Group.
Feel free to put as much information as you like, but it might be good to have a brief introduction to your background, where you are now, specific areas of interest you might have, and on what you are interested in collaborating in the future.
Hi everyone, I am one of the co-chairs of this group. I am a research scientist at the Bureau working in seasonal and marine forecasting and applications. My background is in physical oceanography and ocean modelling. I haven’t done much work with machine learning techniques yet, but I have been following the recent literature on ML-based weather and seasonal forecasting closely. I am particularly interested in ML-based ocean forecasting applications and coastal downscaling methods.
I am looking forward to interacting with everyone and building this community!
Hi. I am not sure if I am a card carrying member of this working group but I am interested. I am an Associate Professor in the School of Maths and Stats at UNSW. My research is predominantly in the oceans role in climate. I have a strong methods focus to my work which dabbles in dynamics, thermodynamics, state estimate/inverse modelling, and machine learning. I like using different data sources from small scale in situ obs to large scale model ensembles. I developed a graduate level course in Environmental Data Science at UNSW which got me more interested in the stats ML world. I am keen to do more and also connect our talented maths/stats focused students to weather and climate problems.
Hi, I am a co-chair of this group and a Senior Research Associate at the Centre of Excellence for 21st Century Weather, based at UNSW Sydney. I have a multidisciplinary background in Computer Science, Applied Mathematics, Environmental Science, and Climate Science. My research focuses on developing machine learning methods for various climate and weather applications, particularly in climate downscaling and drought analysis.
Hi I’m Terry O’Kane,
I’m Senior Principal Research Scientist and Group Leader at CSIRO. I’m an applied mathematician and have been working over the last decade with data science colleagues in Europe and at CSIRO on developing machine learning methods for reduced order modelling, Bayesian Inference, regression learning and entropic AI with applications for climate model emulation and evaluation. I’m the CSIRO representative for the UK MetOffice Partnership committee for ML/AI. I’m really happy to see this initiative develop at ACCESS NRI.
Hi everyone. I am also one of the co-chairs of this group. I am the team leader of the Data Science and Emerging Technologies team in the Research program at the Bureau of Meteorology. My background is in computer science, data science, verification and programming. I am co-author of several machine learning papers. I am the maintainer of the open source verification package “scores” (scores.readthedocs.io).
My current research focus is high-resolution limited-area neural earth system modelling. In the past I have explored machine learning for site-based forecasting and machine learning for global modelling. I am interesting in collaborating with others, building good common tools, and having lots of discussions.
I’m Yue Sun from NCI. I’m a physicist and have been working on SFNO and other deep learning models, such as diffusion and neural ordinary differential equations, in the context of climate and weather studies, in the last 18 months or so. I have a broad interest in DL models that incorporate with physical laws for general science and engineering applications. My current focus is long rollout stability and ensembles
I am looking forward to collaborations with Gadi users that have the same interests in the next 6 months or so.
Hi, I am Abhnil Prasad (Abhi) from UNSW. I am a senior research fellow working on ML applications in weather, climate, and the energy sector. We are currently developing a hybrid climate model by emulating moist physics and radiation in CESM. We have also developed TorchClim, a plugin to integrate Pytorch-based ML models into Fortran-based GCMs.
I am looking forward to working with researchers with similar interests.
Hi all, I’m Cat de Burgh-Day from the Bureau of Meteorology. I’m in the Coupled Modelling Team, and am leading a project to explore, evaluate and develop global ML weather and climate models. Our initial focus has been evaluation of GraphCast and FourCastNetV2 on weather to S2S timescales. We have also done work on developing our own global ML model designed for weather to S2S timescales called SERAS. I am also a member of the WMO Research Board, and though that am co-chair of the AI for Weather Task Team. I am excited to see increased collaboration within Australia through this working group, and am keen to learn all about what everyone else is up to!
Hi all, I am part of the ACCESS coupled model team at CSIRO mostly working on model evaluation. I have some ML experience from a collaboration using deep learning but am still learning.
Thanks, Peter
I am also one of the co-chairs of this group. I’m a Senior Research Scientist at CSIRO in the Geophysical Fluids group. I have a background in the numerical simulation and data-driven stochastic subgrid turbulence modelling in atmospheric, oceanic and engineering flows. I was also one of the co-developers of the Climate Analysis and Forecast Ensemble (CAFE) system, involving data assimilation, parameter estimation and ensemble forecasting methods. Most recently I have been exploiting our CAFE reanalysis to build data-driven yet physics constrained climate emulators. I have also been collaborating with researchers from other fields using machine and statistical learning to understand and predict how the climate influence human factors (e.g. agricultural production, commodity prices, load on healthcare sector).
Looking forward to helping to build a strong and vibrant community.
Hi, Steve Sherwood, Professor at the Climate Change Research Centre UNSW. I work mainly on convective-scale atmospheric processes and how they interact with larger scales. Am hoping ML can help. Have been working with several others in the group on ML downscaling and developing a hybrid climate modelling framework.
I’m Jemma Jeffree, a PhD student at the ANU studying El Niño Southern Oscillation (ENSO). I’ve not yet applied machine learning, but am interested in following the work being done with climate emulators and smaller ML models, and the way they can bridge the gap between statistical methods and physics-based GCMs. I’m also interested in the possibility of ML modelling ENSO with more realistic (or observationally tuned) characteristics than current CGCMs, or using ML to create ENSO ensembles at smaller cost than CGCM ensembles.