Gerard Marcet Carbonell’s research journey in EXPECT began with a fascination for the complex, often unintuitive behaviour of climate systems. After earning an undergraduate degree in physics, he pursued a master’s in modelling for science and engineering. During his studies, a course in climate physics sparked a burning question for him: ‘What physics are we missing in the models?’
This conundrum led him to the Barcelona Supercomputing Center, where he is now a PhD researcher in the Department of Earth Sciences who is working to understand why models fail to fully capture the reality of atmospheric circulation.
What are the gaps in current climate predictions?
‘Summer atmosphere circulation in the Northern hemisphere is not being well captured by current models,’ says Marcet as he highlights uncertainties about future climate patterns. This challenge, along with the promising juncture between climate sciences and physics have been the main catalysts for Gerard’s involvement in the research.
What have you been working on within EXPECT?
‘I’m mostly focusing on climate change drivers, but I’ll soon study how they affect extreme events,’ he says.
Marcet’s work aligns with two key themes of EXPECT:
- Theme 2: Enhancing climate predictions with integrated attribution – Identifying the specific mechanisms driving climate change, such as external forcings (e.g., greenhouse gases, aerosols, and solar radiation) and internal variations within climate systems.
- Theme 3: Unravelling the complex relationships between atmospheric dynamics and extreme weather events, and understanding how these changes affect extreme events to better understand and predict them.
Find more about climate forcing within EXPECT in this read
Gerard Marcet sheds light on the behind-the-scenes work that turns complex climate models into actionable insights. While the heavy computational lifting is done by model developers, researchers like him focus on refining and interpreting data. ‘There’s always post-processing to be done,’ he notes.
Using Python, he extracts key variables from massive datasets, converting raw data into interpretable plots and tables. These visualisations – whether world maps showing rainfall patterns or abstract metrics – allow scientists to analyse climate trends and expand the repository of digested output data.
Have you seen any recent breakthroughs?
‘I’ve been working in an analysis print for LESFMIP not too long ago: it was 4 days of intensive work on this experiment,’ Marcet says. The Large Ensemble Single Forcing Model Intercomparison Project (LESFMIP) is a series of climate simulations designed to isolate specific thermodynamic components of the climate system.
‘Instead of looking at the whole picture, we analyse one set of simulations at a time,’ he adds. By breaking down factors such as incoming and outgoing radiation, volcanic influence, and ozone levels, researchers aim to better understand how each element contributes to climate dynamics and, in turn, improve predictions of atmospheric behaviour.
One very preliminary takeaway from the experiment challenges common assumptions:
‘Greenhouse emissions seem to not be the driving force behind recently documented changes in atmospheric circulation that have been linked to extreme weather events, but rather aerosols are. Despite greenhouse emissions possibly not playing a role in this specific change, the generalised warming associated to these emissions increases the impacts that the extreme heat events have.’
The future of climate models
Marcet’s research focuses on identifying gaps in current models. ‘We always advance by corrections,’ he explains, emphasizing that climate models evolve through iterative improvements. While models alone are not the sole source of climate knowledge, they serve as crucial tools for predicting future tendencies.
Regardless of their limitations, Marcet stresses that strong conclusions about climate trends can still be drawn by integrating models with fundamental physical principles.
Hooked on climate models? Get firsthand insights from the researchers shaping EXPECT and the future of regional climate predictions right here.
We’re on LinkedIn, X, and Bluesky. Follow us there if you haven’t already!