Satellite-Driven Predictions and Projections of Scandinavian Heatwaves (ScanHeat)

Climate models have struggled to represent the persistence and intensity of Scandinavian heatwaves. This limitation is partly due to an incomplete understanding of internal climate variability and the complex drivers behind these events including changes in the North Atlantic Ocean and Arctic sea ice. By identifying and validating satellite-based climate indicators, ScanHeat will make it possible to better predict how and when extreme heat events develop, and how they might evolve in a warmer future climate.

Objectives

ScanHeat will:

• Identify satellite-derived indicators that influence Scandinavian heatwaves.
• Develop a machine-learning-based seasonal forecasting system using these indicators.
• Improve climate projections by validating heatwave indicators in climate models.
• Collaborate with stakeholders to co-develop impact-oriented tools that support climate services in sectors such as health, energy, agriculture, and forestry.

Approach

Researchers will use the EC-Earth climate model, satellite data on sea surface temperature, sea ice, soil moisture, and vegetation, and advanced signal analysis to separate human-driven trends from natural variability. These results will feed into a probabilistic, physics-informed machine-learning forecast system capable of predicting heatwave risks months in advance.

Partnership and disciplines

ScanHeat brings together expertise from three SMHI disciplines: climate modelling, meteorology, hydrology and remote sensing. The project will also include collaboration with Stockholm University. Additionally, the project will engage different nordic stakeholders.