View this email in your browser
Climate Change AI Newsletter
With the impacts of the COVID-19 pandemic being felt globally, we hope that all of you are finding ways to cope and stay healthy. Amid the uncertainties, we’ve found hope in the initiative that people from all walks of life have taken to support one another.

In this newsletter, we’ve assembled opportunities to engage in climate change and machine learning, from across the community. We’ve also included readings and resources that we found interesting.

Do you have opportunities to share or other content you would like to see included in the newsletter? Get in touch at For discussion with fellow readers, follow @ClimateChangeAI on twitter or join our forum.

Workshop at International Conference on Learning Representations

Due to concerns over COVID-19, the Climate Change AI workshop at ICLR 2020 will now be held online on April 26. Registration is open, and has been reduced to $100 ($50 for students). We're excited by this opportunity to exchange and develop ideas, and we invite you to listen to keynotes from experts on climate change and machine learning, pose questions to panelists from across the community, and virtually meet workshop participants from around the world. This is also an opportunity to experiment with remote conferencing, with an eye towards reducing future emissions from travel.
Calls for Submissions

The wider community is hosting several workshops with a strong emphasis on interdisciplinarity and topics related to climate change and machine learning.
  • The workshop on Tangible Information Systems Solutions as a Response to Climate Change at the European Conference on Information Systems will explore efficiency and emissions in energy generation, infrastructure, and consumption. Abstract deadline extended to April 1.
  • The workshop on Data Science in Climate and Climate Impact Research at ETH Zürich will examine conceptual issues underlying the use of data science methods in climate relevant domains, inviting perspectives from across disciplines. Abstract deadline extended to April 6.
  • The workshop Machine Learning Advances Environmental Science at the International Conference on Pattern Recognition is seeking 6 - 10 page papers describing applications of ML to environmental sensor data, which can be used to monitor biodiversity, inform crisis response, or assess GHG emissions, for example. Papers due June 15.
Challenges and Grants

From accelerated science to mapping city climates, a variety of interesting challenges and grant opportunities have been announced.

  • The European Space Agency is providing grants to companies that use earth observation data to enable transitions to green buildings and sustainable built environments. Proposals due April 13.
  • The US Department of Energy has announced grant opportunities on (1) accelerated science and (2) fusion energy, both emphasizing the role of machine learning. Proposals due April 30.
  • National Geographic and Microsoft AI for Earth have opened applications for their AI for Earth Innovation grants program. Projects supporting climate change resilience, adaptation, and mitigation are specifically invited. Proposals due July 22.
  • The Call for Code Global Challenge is now open, with one track on climate change and its impacts. The challenge is open to startups, academics, and enterprises. Submissions due July 31.
  • While registration has closed for the HIDA Datathon on Climate Change, it is still possible to put your name on a waiting list. The challenge will be held in Berlin on November 5 and 6.

Closed-loop optimization of fast-charging protocols for batteries with machine learning
Peter M. Attia, Aditya Grover, Norman Jin, Kristen A. Severson, Todor M. Markov, Yang-Hung Liao, Michael H. Chen, Bryan Cheong, Nicholas Perkins, Zi Yang, Patrick K. Herring, Muratahan Aykol, Stephen J. Harris, Richard D. Braatz, Stefano Ermon, William C. Chueh

Given 10 minutes to charge a lithium-ion battery, what strategy will result in the longest expected battery life? This is a difficult question, because it requires comparing many charging protocols, each of which takes time to experimentally evaluate. The authors address this challenge through Bayesian optimization, which proposes worthwhile charging strategies, and regularized regression, which predicts battery life from a partial experiment. They use these methods in the lab, and find an effective charging strategy in 16 days (instead of 500).

Machine learning and artificial intelligence to aid climate change research and preparedness
Chris Huntingford, Elizabeth Jeffers, Michael Bonsall, Hannah Christensen, Thomas Lees and Hui Yang

The authors summarize the range of ways machine learning is currently used in climate science studies, from earth systems modeling to forecasting climate impacts. They highlight specific climate science gaps which could benefit from a machine learning perspective, including understanding the 2018 UK drought, the warming hiatus, and equation building. They then consider machine learning for climate adaptation, including an overview of the drought forecasting problem.

Up to two billion times acceleration of scientific simulations with deep neural architecture search
Muhammad Kasim, Duncan Watson-Parris, Lucia Deaconu, Sophy Oliver, Peter Hatfield, Dustin Froula, Gianluca Gregori, Matt Jarvis, Samar Khatiwala, Jun Korenaga, Max Topp-Mugglestone, Eleonara Viezzer, Sam Vinko

Many physical simulators can be emulated using machine learning, but effective emulators often require many example simulation runs for training, preventing more widespread adoption. The authors propose to learn good architectural priors for the limited example regime, using neural architecture search. They investigate accuracy of the resulting emulators on 10 simulators, including atmospheric and oceanic cycling models. Further, the acceleration provided by emulators enables sampling-intensive procedures, like uncertainty quantification using MCMC.

Interesting discussions have emerged on the Climate Change AI forum, ranging from advice for specific projects to foundational knowledge for the field. Here are some highlights:

Through the online forum and responses from readers like you, we keep track of student and career opportunities at the intersection of climate change and machine learning.

  • Student research assistantships on data analysis [Potsdam Institute for Climate Impact Research]
  • PhD positions in carbon mapping [University of Copenhagen, Kayrros, LSCE, INRAE]
  • Tenure-track faculty position in Department of Environmental Science, Informatics, and Statistics [Ca’ Foscari University of Venice, April 3]
Industry and Government
Copyright © 2020 Climate Change AI, All rights reserved.

Want to change how you receive these emails?
You can update your preferences or unsubscribe from this list.

Email Marketing Powered by Mailchimp

Images from the Noun Project, by ProSymbols, Nithinan Tatah,, Rudez Studio, and chappara.