Introduction
Cambridge University Press is pleased to be collaborating with Climate Change AI - a global initiative to catalyse impactful work at the intersection of climate change and machine learning – on the following Call for Papers that will lead to a special collection in Environmental Data Science (EDS). This is linked to the Tackling Climate Change with Machine Learning series of workshops that take place at NeurIPS and ICLR conferences. Participants are welcome to submit their work to EDS for the special collection dedicated to the workshop series.
Scope
This Special Collection will focus on the use of artificial intelligence (AI) and machine learning (ML) to help address climate change, encompassing mitigation efforts (reducing greenhouse gas emissions), adaptation measures (preparing for unavoidable consequences), and climate science (our understanding of the climate and future climate predictions).
Specifically, we aim to: (1) showcase high-impact applications of AI/ML to climate change mitigation, adaptation, and climate science, (2) discuss related research directions to which the AI/ML community can contribute, (3) brainstorm mechanisms to scale early academic research to successful, viable deployments, and (4) encourage fruitful collaboration between the AI/ML community and a diverse set of researchers and practitioners from climate change-related fields.
Topics
The Special Collection is expected to feature climate-relevant applications of machine learning to a wide variety of sectors and topics, including:
Climate related topics:
Agriculture and food; behavioral and social science; buildings and transportation; carbon capture and sequestration; cities and urban planning; climate finance; climate justice; climate policy; climate science and modeling; disaster management and relief; earth observations and monitoring; earth science; ecosystems and biodiversity; extreme weather; forestry and land use; health; heavy industry and manufacturing; local and indigenous knowledge systems; materials science and discovery; oceans and marine systems; power and energy systems; public policy; societal adaptation and resilience; supply chains; transportation.
ML-related topics:
Active learning; causal and bayesian methods; classification, regression, and supervised learning; computer vision and remote sensing; data mining; generative modeling; hybrid physical models; interpretable ML; meta and transfer learning; NLP; recommender systems; reinforcement learning and control; time series analysis; uncertainty quantification and robustness; unsupervised and semi-supervised learning.
Metrics/Impact related topics:
As AI/ML areis increasingly used for climate action, the ultimate efficacy and deployability of the proposed methods will depend on the quality of the metrics that are used to develop them. However, there is often a divergence between the metrics used by AI/ML practitioners and the metrics valued in the climate-relevant domains in which AI/ML algorithms are meant to be deployed. For example, while there is much research on safe and robust machine learning (e.g., in the areas of safe reinforcement learning and adversarially robust deep learning), the notions of “safety” and “robustness” used within AI/ML often differ from the notions used within power grids (which stem from, e.g., electrical engineering and control theory); this serves to potentially hamper the deployment of ML within power grid decarbonization workflows.
In addition, as AI/ML methods are deployed for climate action (and more broadly), it is important to understand what their impacts on climate change mitigation, adaptation, and climate equity actually are; however, there is a lack of proper metrics and impact assessment frameworks to evaluate this in practice. For instance, practitioners may want to quantify the greenhouse gas emission savings and increases resulting from not only the computational energy used for the model but also from how it is applied in a particular setting; however such evaluation metrics and tools are only beginning to be developed.
Timetable
- Workshop: Submission deadline: September 18th
- Workshop: Decisions on papers: October 20th
- Workshop takes place: December 9th
- EDS submission deadline: Authors welcome to submit as soon as they are ready (final deadline: February 28th 2023)
Articles will be published as soon as possible after acceptance, in the interest of allowing authors to disseminate their work without unnecessary delay, and added to a curated page for the collection of articles. An editorial reflecting on the insights of the articles will be published as the collection closes.
Please note the following key details, with more information available in the EDS Instructions for Authors:
Article Types: We assume full papers accepted into the workshop will either be submitted to EDS as application or methods papers, with authors of position papers and policy notes for the workshop invited to submit either position papers or perspectives. A full list of article types is here.
Templates: EDS LaTeX and Word templates are available but authors are not required to use these. Authors using the ACM template will need to adapt the article to include:
- Impact Statement: 120 words beneath the abstract describing the significance of the findings in language that can be understood by a wide audience
- Author contributions (using the CRedIT taxonomy as a guide)
- Competing interest statement
- Data availability statement
- Funding statement
See the EDS Instructions for Authors for more details about these statements. Note that we strongly encourage authors to make replication code and data available via open repositories, which should be linked to in the Data Availability Statement. Authors doing so will be awarded Open Data and Open Materials badges on publication.
Submit the paper through the EDS ScholarOne site and select 'Tackling Climate Change with Machine Learning' when prompted to in the question about the special issue in the form.