A scientist from the Karlsruhe Institute of Technology (KIT) in Germany has received a grant from the European Research Council (ERC) to support a project aimed at improving subseasonal weather forecasts while reducing computational requirements, saving energy and costs.
The Advancing Subseasonal PredIctions at Reduced computational Effort (ASPIRE) project will receive up to €1.5m (US$1.6m) in funding annually for up to five years.
KIT meteorologist Dr Julian Quinting utilizes recurring signals in the tropical Pacific that have an important influence on atmospheric circulation in Europe. Additionally, he develops machine learning models to imitate the effects of a high resolution.
Quinting and his future working group are not only aiming to improve the precision of subseasonal forecasts (3-6 weeks) but also to reduce the computing effort, which would save costs and energy and in consequence also reduce greenhouse gas emissions.
Quinting said, “My underlying idea is to make more use of sources in the atmospheric system with high intrinsic predictability. These sources are for instance recurring patterns in the atmosphere that vary on the time scale of two weeks to two months.”
The meteorologist believes recurring signals in the tropical Pacific to be especially promising as these have a major influence on the atmospheric circulation in Europe. These tropical signals however are insufficiently represented in numerical weather prediction models and thus prevent the full exploitation of the underlying intrinsic predictability. In ASPIRE, Quinting plans to improve the representation of the tropical signals using a high spatial resolution in the tropics. Such a high resolution however typically needs more computing power. To avoid this, Quinting and his working group also develop machine learning models that imitate the effects of a high resolution, helping to reduce the computing effort.
“With ASPIRE, we want to showcase the potential of simulations with locally high spatial resolution,” Quinting said. “Ideally, weather services will be able to utilize existing computing power even better.”
If the selected approach proves successful, it could be used for climate research on other components of the atmospheric system that also have a high intrinsic predictability but are incorrectly represented in weather prediction models.
ERC Starting Grant 2022
The ERC funds excellent young scientists that want to start an independent career and establish a working group with starting grants. Every selected project receives up to €1.5m (US$1.6m) annually for up to five years. Under certain circumstances, up to €1m (US$1m) can be requested additionally, for example for equipment or access to infrastructure.
In the 2022 round of calls, the European Research Council awarded starting grants to 408 scientists based in 26 European countries, 81 of them based in Germany. The ERC starting grants have a total funding volume of €636m (US$670m); 2932 applications were entered, resulting in an approval rate of 13.9%.