ML for Earth System Modelling and Analytics workshop 2021

3 May 2021, 14:30 → 4 May 2021, 18:10 Europe/Berlin

DKRZ/HZG/GERICS co-organize a workshop, supported by Helmholtz AI, to facilitate knowledge exchange and discussion about the state of the art and future directions of applying ML methods in Earth System modelling, analytics, and impact research.

To overcome limitations in computing and data analytics related to Earth System science, the uptake of artificial intelligence (AI) and machine learning (ML) methodologies is currently being explored. Multiple initiatives are now emerging to tackle open challenges such as subscale parametrization, detection of patterns and in-situ analysis, adoption of ML for alternative process models, or dedicated fast prediction systems to address specific end-user needs.

Following the first workshop held in February 2020, this second workshop will provide an update on the state of the art in applying and extending AI/ML techniques in topics relevant to Earth System science, from integrating ML with models to deriving new insights from observational data, to extending AI/ML to impact models and approaches. We call for contributions to the sessions, offering presentation and lightning talk slots for those interested in submitting an abstract (abstract submission is now closed). Contributions are welcome from all participants, from updates on ongoing research activities and future plans to technical insights worthwhile to share.

The workshop will take place on the afternoons of May 03/04, 2021, from about 14:30-18:00 CEST. The workshop will be an exclusively online event. The workshop is co-organized by DKRZ, HZG and GERICS with support from Helmholtz AI.

Workshop co-organizers:
Laurens Bouwer (GERICS), Christopher Kadow (DKRZ), Tobias Weigel (DKRZ/Helmholtz AI), Eduardo Zorita (HZG)

Monday, 3 May

14:30 → 14:45 Welcome, logistics

14:45 → 15:15 Keynote 1: Leveraging Artificial Intelligence and Big data across the disaster risk management cycle; opportunities and challenges for the Red Cross

Speaker: Marc van den Homberg (510/Netherlands Red Cross)

vdHomberg_510 DKRZ Hereon GERICS 03052021.pdf

15:15 → 15:20 Short break

15:20 → 16:45 Parallel sessions day 1: Earth System Modelling

Convener: Christopher Kadow (DKRZ)

15:20 An ML Perspective on Closed Loop Tuning in Earth Science Simulators

Speaker: David Greenberg (Hereon)

15:35 A machine learning-based air quality forecast system for Pacific Northwest

Speaker: Kai Fan (CASUS/HZDR)

15:50 Near-surface temperature forecasting by deep learning

Speaker: Bing Gong (Juelich Supercomputing Center)

2021-04-24_AI_Earth_workshop_Bing.pdf

16:05 Using machine learning techniques to classify ENSO events

Speaker: Nicola Maher (U. Colorado, Boulder/MPI for Meteorology)

16:20 The Frontiers of Deep Learning for the Earth System Sciences

Speaker: David Hall (NVIDIA)

15:20 → 16:45 Parallel sessions day 1: Extreme Events and Impacts

Convener: Laurens Bouwer (GERICS)

15:20 Machine learning model of the plasmasphere to forecast satellite charging caused by solar storms

Speaker: Stefano Bianco (GFZ Potsdam)

StefanoBiancoMaterial.pdf

15:35 Denoising seismic data using a ResNeXt-50-based convolutional autoencoder

Speaker: Jan Walda (University of Hamburg)

15:50 Artificial Intelligence for flood analysis: first results from the AI4Flood project

Speaker: Shagun Garg (GFZ Potsdam)

16:05 Automated Damage Assessment

Speaker: Jacopo Margutti (Netherlands Red Cross)

Margutti_ML_for_Earth_03-05-2021.pdf

16:20 Supervised Machine Learning to investigate Heat Waves and Myocardial Infarctions in Augsburg, Germany

Speaker: Lennart Marien (GERICS)

16:45 → 17:00 Break

17:00 → 17:15 Neural Earth System Modelling

Speaker: Christopher Irrgang (GFZ Potsdam)

17:15 → 18:00 Discussion (co-chaired by C. Kadow & L. Bouwer)

Tuesday, 4 May

14:30 → 15:10 Lightning talks

Convener: Tobias Weigel (DKRZ)

15:10 → 15:40 Lightning talks: Lightning talks interaction session

15:40 → 15:45 Short break

15:45 → 17:15 Parallel sessions day 2: Earth System Modelling

Conveners: Christopher Kadow (DKRZ), Eduardo Zorita (Hereon)

15:45 3D bias correction with deep learning in the Integrated Forecasting System

Speaker: Thorsten Kurth (NVIDIA)

Kurth_ECMWF Data Assimilation Bias correction.pdf

16:00 Building an Earth System Model emulator for local monthly temperature

Speaker: Shruthi Nath (Climate Analytics/ETH Zürich)

16:15 Machine learning based post-processing of 2-m air temperature model output – a multi-model approach

Speaker: Gabriel Stachura (Institute of Meteorology and Water Management (Poland))

Stachura_ML_workshop.pdf

16:30 Lossy Compression of Climate Data using Convolutional Autoencoders

Speaker: Silke Donayre Holtz (KIT)

Climate_Data_Compression_Donayre.pdf

16:45 Building a versioned pipeline for ML-based quality control of environmental sensor data

Speaker: Julius Polz (KIT)

15:45 → 17:15 Parallel sessions day 2: Remote sensing and marine ecosystems

Convener: Tobias Weigel (DKRZ)

15:45 Quantifying the impact of bedrock uncertainty on ice sheet model simulations by Gaussian Process modelling

Speaker: Andreas Wernecke (MPI for Meteorology)

Wernecke_ML_bed.pdf

16:00 Remote Sensing of Global Ocean Wind Speed using GNSS Reflectometry and AI

Speaker: Milad Asgarimehr (GFZ Potsdam)

16:15 Approximation of a marine ecosystem model using artificial neural networks

Speaker: Markus Pfeil (Kiel University)

16:30 Artificial Intelligence for Cold Regions (AI-CORE) - a Pilot to bridge Data Analytics and Infrastructure Development

Speaker: Long Duc Phan (AWI)

LongPhan_AICore_DKRZ-workshop.pdf

17:15 → 17:30 Break

17:30 → 18:00 Keynote 2: A mile wide and a pixel deep: Integrating computer vision and satellite imagery for coupled-natural human system modeling

Speaker: Kelly Caylor (UCSB)

Caylor_Helmholtz ML AI Keynote.pdf

18:00 → 18:10 Wrap-up