EARTHLAB — Earth Data Analytics Research Network

About the Programme

EARTHLAB brings together early-career geoscientists at Cambridge for a week of intensive, hands-on training in modern climate data science methods. Over a 5-day period, participants will learn cutting-edge quantitative methods and apply them directly to their own research through collaborative hackathons, peer-led problem-solving, and specialist lectures.

The goal is to cultivate a lasting community within the Cambridge ecosystem that fosters interdepartmental collaboration, shares analytical capacity, and produces tangible tools, datasets, and publications.

12 Participants

5 Days

Free To attend

AI Claude Pro included

Interdisciplinary

Bringing together researchers from climate science, environmental science, and beyond

Collaborative

Work in teams towards common research goals, building a network that extends far beyond the programme

Hands-On

Interactive lectures, live coding sessions, and real-world data challenges using your own research data

Future-Facing

Learn to integrate AI coding assistants into your research pipeline — from data cleaning to analysis and visualisation

Programme Structure

An intensive one-week programme running 15–19 June 2026, mixing lectures with hands-on collaborative work.

Welcome & Presentations Lecture Hackathon Final Presentations

Day 1 – AM

Welcome, Research Pitches & Team Formation

Introduction to EARTHLAB, participant presentations of research interests, and team formation around shared goals and complementary skills.

Day 1 – PM

Lecture: AI Pipelines for Research

Hands-on introduction to AI coding assistants and automated workflows for research—setting up collaborative tools and pipelines that teams will use throughout the programme.

Day 2 – AM

Lecture: Bayesian Modelling

Interactive lecture on Bayesian methods for geoscientific data. Hands-on practical applying probabilistic frameworks to real climate datasets.

Day 2 – PM

Collaborative Hackathon I

Team-based work applying Bayesian approaches to participant projects. Peer support and problem-solving.

Day 3 – AM

Lecture: Machine Learning for Climate Data

ML techniques tailored to climate and environmental applications. From dimensionality reduction to neural networks for spatiotemporal data.

Day 3 – PM

Collaborative Hackathon II

Teams continue building on their projects, integrating ML methods. Mid-programme check-in and progress sharing.

Day 4 – AM

Collaborative Hackathon III

Full-day deep work on team projects, combining all methods learned.

Day 4 – PM

Collaborative Hackathon III (cont.)

Continued project development with mentor support and peer feedback.

Day 5 – AM

Final Hackathon & Polish

Last sprint to refine projects, prepare visualisations, and work towards publishable results.

Day 5 – PM

Presentations & Next Steps

Teams present their work, discuss potential pathways to publication, and plan ongoing collaborations.

What You'll Learn

AI-Powered Workflows

Using AI coding assistants to build research pipelines, automate data processing, and accelerate scientific discovery.

Bayesian Modelling

Probabilistic frameworks for uncertainty quantification, hierarchical models, and inference techniques applied to geoscientific data.

Machine Learning

Supervised and unsupervised methods for climate data—from random forests and gradient boosting to deep learning for spatiotemporal patterns.

Climate Data Analysis

Timeseries analysis, spatial statistics, multi-model ensemble techniques, and working with large-scale observational and reanalysis datasets.

Global warming projections from CESM2 showing temperature anomalies at 1900, 1950, 2010, and 2060

Who Should Apply

Brain and circuit illustration representing ML and AI

EARTHLAB is primarily designed for PhD students and early-career researchers (including postdocs and research staff) working on climate and environmental problems at the University of Cambridge.

Applications from MPhil and Part III students with a well-defined research project in climate or environmental science are welcome and will be considered subject to availability.

We welcome applicants from across the University—Geography, Earth Sciences, Engineering, Plant Sciences, and any other department working on Earth, climate, and environmental data.

This is an intensive data science programme. While no prior experience with Bayesian statistics, machine learning, or AI tools is required, we expect participants to have strong computing skills and a genuine interest in tackling quantitative climate and environmental problems. The programme is free of charge, fully funded by the Department of Geography's SRIF fund.

Accepted participants will receive a complimentary 3-month Claude Pro subscription to continue building AI-powered research workflows beyond the programme.

Frequently Asked Questions

Do I need prior experience with Bayesian statistics or machine learning?

No. EARTHLAB is designed to be accessible to researchers at all levels of quantitative experience. The interactive lectures will introduce concepts from the ground up, and the collaborative format means you'll learn alongside peers with complementary skills.

What should I bring?

Bring your laptop, your own research data or a specific problem you'd like to tackle, and an open mind. We'll provide a 3-month Claude Pro subscription and all the collaborative infrastructure you need.

Is there a cost to participate?

No. EARTHLAB is fully funded by the Department of Geography's SRIF fund, with additional support from the Centre for Climate Repair and the Cambridge Centre for Climate Science. Participation and AI tools are all provided at no cost.

What happens after EARTHLAB?

EARTHLAB is designed to build a lasting network. We aim to support ongoing collaborations and co-authored publications.

What will I leave with?

Participants leave EARTHLAB with new analytical skills in Bayesian modelling and machine learning, reusable code and workflows they can apply immediately to their research, a network of collaborators across Cambridge departments, and a 3-month Claude Pro subscription to continue building AI-powered research pipelines.

How are teams formed?

On Day 1, participants will pitch their research projects and interests. Teams will form organically around shared goals and complementary skills, with guidance from the organisers to ensure productive collaborations.

How many places are available?

We expect to welcome 12 participants for this inaugural edition of EARTHLAB.

What programming language will be used?

All sessions use Python in Jupyter notebooks. We will also introduce AI coding assistants to accelerate your workflow. No advanced Python experience is required — we will cover the essentials.

Do I need to attend all 5 days?

We strongly encourage full attendance as the programme is designed as a cohesive experience, with each day building on the previous. However, if you have unavoidable commitments, please mention this in your application.

Where does it take place?

EARTHLAB is hosted by the Department of Geography at the University of Cambridge. The exact venue and room details will be confirmed closer to the date.

Can I apply if I'm not based at Cambridge?

EARTHLAB is primarily aimed at researchers within the University of Cambridge, including affiliated institutions such as the British Antarctic Survey. If you are from outside Cambridge but have a strong case for participation, we encourage you to apply.

Organisers

Dr Francesco Muschitiello

Associate Professor in Physical Geography

Department of Geography, University of Cambridge

Palaeoclimatologist specialising in rapid climate transitions, proxy data from marine sediments and ice cores, and quantitative methods for Earth system science.

Profile

Dr Alice Cicirello

University Assistant Professor in Applied Mechanics

Department of Engineering, University of Cambridge

Expert in uncertainty quantification, physics-enhanced machine learning, and data-driven methods for engineering applications.

Profile

Supported By

Apply

Applications are now open. To apply, please fill in the form below with:

A brief description of your current research
What you hope to achieve by joining EARTHLAB
What datasets or research questions you plan to bring

Open Application Form

Applications will be reviewed on a rolling basis.
For questions, contact fm476@cam.ac.uk.