As renewable energy penetration grows, the variability and uncertainty of weather are becoming first-order drivers of electricity grid reliability and energy market volatility. Prolonged wind lulls, sudden cloudiness, jellyfish affecting the cooling of nuclear power plants, or compound multi-hazard events can challenge both short-term operations and long-term planning. Traditional deterministic planning and rule-of-thumb margins are no longer sufficient, especially as energy systems become more localised and spatially heterogeneous. Advanced probabilistic and high-resolution approaches are needed to quantify risk, guide investment, and design resilient energy systems.

This session focuses on probabilistic methods that bridge geoscience and power-system analysis, leveraging atmospheric and hydrological data, statistical extreme value theory, and modern machine learning to improve decision-making in renewable-dominated systems. We welcome contributions that advance this interface through
- Probabilistic resource adequacy and extremes, including methods for Loss of Load Expectation (LOLE), Expected Energy Not Served (EENS), Effective Load Carrying Capability (ELCC), compounding weather-driven events, energy-hub design and structural congestion studies.
- Spatio-temporal probabilistic forecasting, including ensemble methods, post-processing, generative models, and calibration of prediction intervals to support operational flexibility and reserve management.
- Geostatistics and data fusion, including combined satellite products, reanalysis data, combined spatial- and point-based observations, and the propagation of weather uncertainty to energy system operation and planning.
- Open tools and reproducibility, including frameworks like PyPSA, Calliope, open datasets and standardised benchmarks.
The goal of this session is to foster a shared methodological language for risk-aware planning and operation, connecting geoscientists, energy researchers, and practitioners to advance probabilistic methods for the energy transition.