Climate AI Nordics

Climate AI Nordics

Nordic network for collaborations around tackling climate change using AI and machine learning.

Will you join the #AIWeatherQuest?

Call for Climate AI Nordics to apply our expertise to a global forecasting competition: join the ECMWF AI Weather Quest

The ECMWF AI Weather Quest, a global machine learning competition led by the European Centre for Medium-Range Weather Forecasts (ECMWF) and endorsed by the World Meteorological Organization (WMO), is now open. It challenges participants to develop AI-driven probabilistic forecasts for the sub-seasonal range — specifically for days 19–25 and 26–31 ahead, a critical window for early warning and operational climate decisions.

ECMWF is especially eager to see participation from the Climate AI Nordics community, which is at the forefront of using AI to tackle climate challenges. The AI Weather Quest is a unique opportunity to contribute to cutting-edge forecasting research while connecting with a global network of scientists and practitioners.

What participants will forecast

Teams are invited to submit probabilistic weekly forecasts for:

  • Near-surface air temperature
  • Precipitation
  • Mean sea level pressure

Forecasts may target one or both lead times:

  • Days 19–25
  • Days 26–31

Participants can work with any dataset, use any programming language, and apply any AI/ML methodology. Forecasts are self-evaluated using ECMWF’s tools before the competitive scoring phase begins later in the year.

Competition timeline

Initial Training Phase (March–August 2025): Participants explore datasets, tools, and evaluation methods. The Testing JJA Period (June–August) allows teams to submit forecasts and self-evaluate performance in a non-competitive setting.

Competitive Phase (from September 2025): Teams submit real-time forecasts that will be scored and ranked using ECMWF’s evaluation framework. Public results will showcase the top-performing approaches.

How to get involved

The competition is open to individuals and teams from research institutions, private companies, and public bodies. To help participants prepare, ECMWF is hosting two technical webinars on 7 May 2025 that will explain how to participate in the Testing Period:

🔗 Explore the Quest and register here: https://aiweatherquest.ecmwf.int

Published:

Categories:

You May Also Enjoy

Efficient and precise annotation of local structures in data

Event date:

Webinar with John Martinsson, RISE and Lund University. Machine learning models now help scientists analyze vast datasets across every branch of science. These models typically improve with more data and larger architectures, mainly through supervised learning. Both training and evaluation therefore rely on labeled datasets. A main challenge is scaling the data labeling effort to the volumes required, because it is costly and label quality can vary. Methods that deliver inexpensive yet accurate labels are therefore essential. This talk examines how to lower annotation cost and increase label quality when labeling local structures in data—for example, a local structure can be a sound event in an audio recording. By detecting the boundaries of such structures automatically, we let annotators focus on supplying concise textual descriptions for the content within those boundaries. In this setting we analyze a widely used labeling method for audio where fixed and equal length audio segments are labeled with presence or absence of an event class. We benchmark it against an oracle method that defines an upper bound, and propose adaptive labeling techniques that achieve higher‑quality labels for the studied datasets at a lower cost.

SatML for large-scale above-ground biomass estimation

Event date:

Webinar with Ghjulia Sialelli, ETH Zurich. The combination of remote sensing and machine learning has made it possible to map forest properties at an unprecedented scale and resolution. In this presentation, I will focus on the application of deep learning techniques to estimate above-ground biomass (AGB), a key metric for tracking forest carbon and ecosystem dynamics. I will begin by introducing our recently published, machine-learning-ready dataset. It features high-resolution (10m) multi-modal satellite imagery, paired with AGB reference values from NASA’s Global Ecosystem Dynamics Investigation (GEDI) mission. Key aspects include the carefully selected geographic coverage, thoughtful integration of diverse satellite data sources, and the establishment of performance baselines using standard deep learning models. Next, I will describe our ongoing efforts to build on said baselines. This was done both through feature and model engineering. I will also mention some promising yet unsuccessful approaches, highlighting some key challenges of the task at hand. Finally, I will discuss future directions, including incorporating uncertainty estimation and exploring the potential for generating a global above-ground biomass map.

Climate AI Nordics Newsletter, March 2025

Published:

Welcome to the Climate Change AI Nordics Newsletter, March 2025! Read about recent and coming seminars, workshops, and a critical perspective of environmentally sustainable AI.