When10–14 May 2027
WhereCampus Ecotekne, Lecce
Submission deadline5 February 2027

About CI 2027

The 16th International Conference on Climate Informatics will take place from 10 to 14 May 2027 in Lecce, Italy, at the Campus Ecotekne of the University of Salento. CI 2027 is hosted by the University of Salento and co-organized with the Euro-Mediterranean Center on Climate Change, Foundation (CMCC).

Climate Informatics is an open community working at the intersection of climate science with statistics, machine learning and data mining. Since 2011 its annual conference has brought together researchers across these fields to share new methods, datasets and results, and to start new collaborations.

Important dates

30 November 2026 Abstract & paper submission opens
5 February 2027 Abstract & paper submission deadline
5 March 2027 Notification of results
2 April 2027 Camera-ready copy of full papers
9 April 2027 Registration deadline
10–14 May 2027 Conference

Deadlines are at 23:59 Anywhere on Earth (AoE) unless stated otherwise.

Call for papers

We invite contributions across the full intersection of machine learning, data science and climate research: work that advances our understanding of the Earth system, supports climate adaptation and mitigation, or makes climate data and models more transparent and accessible.

Topics of interest

  • Knowledge-guided and hybrid Earth-system modelling
  • Prediction, nowcasting, downscaling and extremes
  • Data assimilation, reanalysis and remote sensing
  • Causal discovery, interpretability, calibration and uncertainty quantification
  • Climate impacts, adaptation/mitigation decision support and policy
  • Reusable datasets, benchmarks and evaluation tools
  • Principled positions on ethics, reproducibility and open science

What reviewers look for

Submissions should be clearly relevant to climate science, compare against appropriate baselines and include a rigorous evaluation that addresses uncertainty, robustness and limitations. Authors are strongly encouraged to share code, data and other artifacts so that results can be reproduced.

Submission tracks

There are two tracks, full papers and extended abstracts. Both are due on 5 February 2027 (23:59 AoE) and are submitted through Microsoft CMT (see Submission guidelines).

Full papers

  • Original, unpublished work of up to 8 pages, not counting acknowledgements, references and appendices or supplementary material.
  • Prepared with the Environmental Data Science (EDS) LaTeX or Word templates, available from Cambridge University Press and Overleaf.
  • Accepted full papers are considered for an oral or poster presentation, and accepted papers that are presented at the conference are published in Environmental Data Science.

Extended abstracts

  • At least 300 words and at most 2 pages in total, including figures and references. The work must be original; preliminary or ongoing research is welcome.
  • Must include at least one figure or schematic that visually summarises the research.
  • Suggested structure: Introduction and objectives, Methods and data, Preliminary results and/or expected outcomes (including implications and limitations).
  • Use the CI 2027 template: Word (.docx) · LaTeX (.zip)
  • Accepted extended abstracts are considered for a poster presentation.

Paper categories

When submitting, choose the single category that best describes the main contribution of your work. Categories are used to assign reviewers and follow the article types of Environmental Data Science. They apply to both tracks, except position papers, which are accepted as full papers only.

  • Application papers: research progress on, or a solution to, a real-world problem in an environmental field, made possible by data science (for example, using AI to better understand climate applications or to improve forecasting tools).
  • Methods papers: new data-science methodology motivated by an environmental problem, usually demonstrated on one or more climate applications.
  • Data papers: the creation and usefulness of reusable environmental datasets or benchmarks. The data must be available in a publicly accessible repository.
  • Position papers (full papers only): well-argued views on the adoption of AI and data science, or on obstacles to it, or on environmental data issues such as ethics, law, policy, standards and services.

Review process

  • Reviewing is double-blind. Manuscripts and supplementary material must not reveal the authors’ identity, directly or indirectly; submissions that do may be desk-rejected.
  • Posting related work on arXiv is allowed. If you cite your own preprint, refer to it in the third person and avoid wording that reveals authorship.
  • Reviewers’ identities are not disclosed to authors.

Submission portal opens 30 November 2026

Submission guidelines

Full papers and extended abstracts follow the same submission and review process. See the Call for papers for formats and page limits.

Deadline

Full papers and extended abstracts are due on 5 February 2027 (23:59 AoE). Supplementary material for full papers must be submitted by the same date.

Submission system

Submissions are handled through Microsoft CMT (Conference Management Toolkit). Make sure you submit to the correct track and select the paper category that best fits your work.

Anonymization

Reviewing is double-blind, so the initial submission must not include author names or affiliations. Remove or rephrase anything that could identify the authors, such as acknowledgements, self-citations and grant numbers. These can be restored in the camera-ready version after acceptance.

Originality and reuse of text

  • Submissions must present new, original work that has not been published elsewhere and is not under review at another venue at the same time. A paper counts as published if it appeared in an archived, peer-reviewed venue that people who did not attend can access.
  • Concurrent submissions will be rejected without review.
  • Reusing previously published text verbatim is not allowed. Previously published images may be reused only if they are cited and the necessary rights have been obtained.
  • Allowed exceptions: preprints on non-archival repositories such as arXiv, citations of theses, and new versions of earlier non-archival work (such as posters or workshop papers) that add substantial new material and cite the earlier version.

Supplementary material

Supplementary material such as code, data, evaluation protocols, additional results and figures is encouraged. Report the key parameters of your algorithms, use open data where possible and, if you rely on proprietary or external data, share a version of it if you can. The main submission must stand on its own: reviewers may look at supplementary material but are not required to.

Use of generative AI

Authors may use generative AI tools when preparing their manuscripts. AI tools cannot be listed as authors, and the human authors remain fully responsible for the content, including avoiding plagiarism, misrepresentation and falsification.

Open science

Authors are encouraged to follow open-science practices and to accompany their submission with a project repository (for example on GitHub). The Turing Way explains how to set up a repository, choose a licence and obtain a DOI. Authors of full papers should also follow the Cambridge University Press Research Transparency Policy on sharing data and code.

CMT Acknowledgment. The Microsoft CMT service was used for managing the peer-reviewing process for this conference. This service was provided for free by Microsoft and they bore all expenses, including costs for Azure cloud services as well as for software development and support.

Author guidelines

Climate Informatics welcomes high-quality, novel contributions that bridge climate and weather research with data science and advance data-driven methods in the atmospheric and climate sciences. Submissions can be full papers or extended abstracts; see the Call for papers and the Submission guidelines for details.

Formatting requirements

  • Full papers use the Environmental Data Science templates; extended abstracts use the CI 2027 template (Word · LaTeX).
  • All files, including supplementary material, must be in PDF format and written in English.
  • Figures and tables must be readable both in colour and in greyscale. Colour-blind-safe palettes are recommended for all coloured figures.

Submission requirements

  • Submissions are made through the conference CMT portal. Submission portal: link available from 30 November 2026.
  • The corresponding author must make sure that all co-authors approve the results presented and the final manuscript.

Review and publication

  • Each submission is reviewed double-blind by at least two independent reviewers.
  • Remove all identifying information (names, affiliations, acknowledgements) before submitting: submissions that are not anonymized will be desk-rejected.
  • Accepted full papers that are presented at the conference will be published in a special collection of Environmental Data Science. Instructions for this step will be sent after the conference.

Registration and attendance

At least one author of each accepted full paper or extended abstract must register for the conference and present the work.

Good practice

  • Climate Informatics brings together climate scientists, meteorologists, data scientists and policy researchers: write for an interdisciplinary audience.
  • Follow reproducible-research practices: include a data availability statement, links to code repositories and, where possible, to open datasets.
  • Authors from under-represented backgrounds and regions are strongly encouraged to submit.

For questions about submissions or formatting, write to climate.informatics.workshop@gmail.com.

Presenter guidelines

Talks

  • Timing: standard talks last 10–12 minutes, lightning talks 6–8 minutes, both including questions. Plan for roughly one slide per minute and stay within your slot.
  • Slides and videos: upload your slides well in advance and test any videos on the presentation system before your session.
  • Know your audience: the room mixes computer scientists and Earth scientists, so assume as little background knowledge as possible.
  • Keep the maths light: define every term and show only what is needed to understand your results.
  • Readable figures: labels, titles and text should be large enough to read from the back of the room.
  • Little text: keep text on slides to a minimum, so that the audience listens rather than reads.
  • Practise: rehearse in front of a friendly audience to get feedback on delivery and timing.
  • Enjoy it: the audience wants you to succeed.

Posters

  • Format: A0 portrait (841 × 1189 mm), to fit the poster boards.
  • The 3-second rule: use a clear, bold title and one central take-away, so that someone walking past can grasp your main finding in about 3 seconds from 2 metres away.
  • White space: aim for about 40% visual content (figures, charts, maps) and leave open space to guide the eye.
  • Visual hierarchy: put your most important result in the centre and use headings to lead readers from the introduction (top left) to the conclusions (bottom right).
  • Interdisciplinary clarity: define acronyms and avoid niche jargon. Earth-science posters can follow the EGU poster guidelines (minimum font size 16 pt); ML/CS posters can use an evidence-first layout, as recommended by the MIT Communication Lab, and avoid heavy maths.
  • Legibility: text and figure legends should be readable from 1.5–2 metres, with high contrast between text and background.
  • Elevator pitch: prepare a 20-second summary for quick visitors and a 3-minute walk-through for interested experts.
  • QR code: add a QR code linking to your full paper, code repository or supplementary videos.

For technical questions about your presentation, contact the organizing committee at climate.informatics.workshop@gmail.com.

Programme

The detailed programme will be published.

Hackathon

As in previous editions, CI 2027 will include a hackathon where participants work in teams on climate data challenges, combining domain expertise with data-science methods.

  • Date: TBA
  • Challenges & data: TBA
  • Sign-up: TBA

Registration

The registration deadline is 9 April 2027. Opening date, fees and what is included will be announced here.

Venue & travel

Campus Ecotekne, University of Salento

Strada Provinciale Lecce–Monteroni, 73100 Lecce, Italy. The campus is about 6 km from Lecce city centre.

Open in OpenStreetMap · Open in Google Maps

By plane

Brindisi airport (BDS) is the closest, about 40 km from Lecce. Direct buses run from the airport to Lecce; alternatively take the shuttle bus to Brindisi railway station and a train to Lecce (about 30 minutes).

Bari airport (BRI) is about 150 km away: take the train link to Bari Centrale, then a train to Lecce.

By train

Lecce is the southern end of the Adriatic railway line (Bologna – Bari – Lecce) and is also reachable from Rome. Check timetables on Trenitalia. Most hotels in the centre are within walking distance of the station.

From Lecce to the campus

City bus line S13 connects the centre with the campus (stops “Ecotekne”). Taxis are also available.

Accommodation

We recommend staying in Lecce’s historic centre. A list of suggested hotels and any agreed rates: TBA.

Committee

  • General co-chairs: Soheil Shayegh (CMCC), Annalisa Bracco (CMCC)
  • Programme committee co-chairs: Italo Epicoco (University of Salento), Viviana Acquaviva (CUNY City Tech)
  • Local chairs: Italo Epicoco, Massimo Cafaro, Marco Pulimeno (University of Salento)
  • Local organizing committee: Enrica Quartini (CMCC)
  • Hackathon: Julie Keisler (chair), Tobias Finn (ENPC, co-chair)
  • Web and communication: Emanuele Mele (University of Salento)
  • Steering committee liaison: Claire Monteleoni (Inria)

The full programme committee will be announced soon.

Contact

For questions about CI 2027, write to climate.informatics.workshop@gmail.com.

To receive announcements about the call for papers and registration, join the Climate Informatics mailing list.