Copy
View this email in your browser
Climate Change AI Newsletter

Happy New Year from Climate Change AI! A year ago, a group of us mobilized to help the AI community use their skills to tackle climate change. It’s been great to see the response. This year, more than ever, rapid climate action is needed - both from the AI community and across society. 

In this newsletter, the call for submissions for our workshop at ICLR 2020 (including a new mentorship program to help you build your submissions!), interesting readings from our workshop at NeurIPS, and upcoming research opportunities and open positions.

Do you have opportunities to share or other content you would like to see included in the newsletter? Get in touch at
info@climatechange.ai.

Announcements
 

Climate Change AI workshop at ICLR
Submissions are open for our ICLR 2020 workshop “Tackling Climate Change with Machine Learning”. We also strongly encourage authors to consider applying for our mentorship program (applications due Jan 14), which will pair people with mentors having complementary expertise.


Climate Change AI conference track at AMLD
Applied Machine Learning Days (AMLD) at the EPFL in Lausanne, Switzerland is coming up this month with our track “
AI & Climate Change” on January 27 and 28. Registration is still open; register here.


Past workshop at NeurIPS 2019
Watch recordings and read papers from our NeurIPS workshop “Tackling Climate Change with Machine Learning” held on December 14, 2019. Pictures below.

Opportunities
 

Workshop "Data Science in Climate and Climate Impact Research” on conceptual issues, challenges, and opportunities at ETH Zürich, August 20 & 21, 2020. Submit abstracts by March 15 here.

Special issue on Data Innovation in Industrial Ecology by the Journal of Industrial Ecology. Submit papers
here by June 30.

Special issue on Energy Informatics by the BISE Journal. Submit papers
here by July 1.

The U.S. Department of Energy’s Solar Energy Technology Office has released a notice of intent for an upcoming funding opportunity. It includes a potential topic on “Artificial Intelligence (AI) Applications in Solar Energy with Emphasis on Machine Learning.” The full announcement is expected this month. Details
here.

Jobs
 

Yoshua Bengio is currently recruiting a postdoc for the visualizing climate change project at the Mila - Quebec AI Institute in Montreal. The ideal candidate would have a strong background in machine learning and deep learning, specifically in computer vision and generative models, with a clear and strong interest in the environment and climate change in particular. Please apply using the Mila postdoc application form, specifying your interest for this project in the application. 

The University of Pennsylvania is offering a postdoctoral research fellow position in Energy Analytics and Machine Learning. Details
here.

RFF-CMCC European Institute on Economics and the Environment (EIEE) is looking for a researcher to work on the European Research Council (ERC) Starting Grant 2D4D, “Disruptive Digitalization for Decarbonization,” in Milan, Italy. A variety of academic backgrounds will be considered for this position, including, but not limited to, empirical economists, innovation system analysts, economic geographers, and public policy analysts. Apply
here by March 31.

Several research positions with the Alan Turing Institute, the UK’s national institute for data science and artificial intelligence. The Data Science for Science programme is looking to hire two Research Fellows to work on research that applies data science and AI methods to environmental science (apply
here). The Digital Twins of Built Environment group in the Data-Centric Engineering Programme is looking to hire in the area of uncertainty quantification and inference for energy models of built environments (apply here). See also the related forum post.

The National Renewable Energy Laboratory (NREL) is hiring for a number of academic and non-academic positions involving machine learning, for example,
this position. For more information on all the positions available, please refer to NREL's careers website.

Tesla is looking for a Software Engineer in Palo Alto, CA, to help develop and maintain the high-level control systems for Megapacks, Powerwalls, Virtual Power Plants, and microgrids. Apply
here.

Lumina Decision Systems is seeking two energy and climate-related analysts (Senior Energy Consultant and Associate Analyst). Lumina is moving in the direction of incorporating ML into their modeling. Details
here.


From the discussion forum

 

Several positions for software engineers, PhD students, and postdocs at the Universitat de València, Spain, with the Image and Signal Processing (ISP) group. Research topics include: regression, causality & information theory, Earth observation data analysis, physics-aware machine learning, generative modeling, explainable AI and feature ranking, and anomaly detection with applications in the Earth and climate sciences. See the post on the forum and details on the group site.

15 PhD studentships to work with leading climate and machine learning scientists across Europe at the University of Oxford to tackle the issue of aerosol-cloud interactions. Apply here by February 3, 2020 and see the related forum post.

Readings: Best paper awards from the CCAI NeurIPS workshop

We recognized three best paper award winners at our NeurIPS workshop on December 14, 2019. Abstracts for these papers are below; videos for these and other spotlight talks can be found on the
workshop website.
 

Cumulo: A Dataset for Learning Cloud Classes 
Valentina Zantedeschi, Fabrizio Falasca, Alyson Douglas, Richard Strange, Matt Kusner, Duncan Watson-Parris

One of the greatest sources of uncertainty in future climate projections comes from limitations in modelling clouds and in understanding how different cloud types interact with the climate system. A key first step in reducing this uncertainty is to accurately classify cloud types at high spatial and temporal resolution. In this paper, we introduce Cumulo, a benchmark dataset for training and evaluating global cloud classification models. It consists of one year of 1km resolution MODIS hyperspectral imagery merged with pixel-width ‘tracks' of CloudSat cloud labels. Bringing these complementary datasets together is a crucial first step, enabling the Machine-Learning community to develop innovative new techniques which could greatly benefit the climate community. To showcase Cumulo, we provide baseline performance analysis using an invertible flow generative model (IResNet), which further allows us to discover new sub-classes for a given cloud class by exploring the latent space.


Machine learning identifies the most valuable synthesis conditions for next-generation photovoltaics
Felipe Oviedo, Zekun Ren

Terawatts of next-generation photovoltaics (PV) are necessary to mitigate climate change. The traditional R&D paradigm leads to high efficiency / high variability solar cells, limiting industrial scaling of novel PV materials. In this work, we propose a machine learning approach for early-stage optimization of solar cells, by combining a physics-informed deep autoencoder and a manufacturing-relevant Bayesian optimization objective. This framework allows to: 1) Co-optimize solar cell performance and variability under techno-economic revenue constraints, and 2) Infer the effect of process conditions over key latent physical properties. We test our approach by synthesizing 135 perovskite solar cells, and finding the optimal points under various techno-economic assumptions. 



Helping Reduce Environmental Impact of Aviation with Machine Learning 
Ashish Kapoor

Commercial aviation is one of the biggest contributors towards climate change. We propose to reduce environmental impact of aviation by considering solutions that would reduce the flight time. Specifically, we first consider improving winds aloft forecast so that flight planners could use better information to find routes that are efficient. Secondly, we propose an aircraft routing method that seeks to find the fastest route to the destination by considering uncertainty in the wind forecasts and then optimally trading-off between exploration and exploitation. Both these ideas were previously published in
here and here and contain further technical details.

Twitter
Website
LinkedIn
Facebook
Copyright © 2020 Climate Change AI, All rights reserved.


Dates and deadlines reported in this newsletter may change, and CCAI is not responsible for any inadvertent inaccuracies.

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