In the first phase of natESM, we successfully introduced a novel service for the national Earth system modeling (ESM) community—short, intensive collaborations (sprints) focused on preparing individual simulation components for future supercomputers.

In Phase 2 (September 2025 to August 2029), we aim to expand on this foundation by integrating advanced Artificial Intelligence (AI) methods, particularly machine learning (ML), and by providing access to well-defined model configurations, developed and maintained by the community. This phase will focus on further developing non-fluid components like biogeochemistry, land processes, atmospheric chemistry, and land ice, enhancing data handling and analysis, and improving collaboration across the ESM community. 

Our goal is to build a world-leading, sustainable national ESM system that serves a broad spectrum of users, from education to cutting-edge research and operational applications. 

Key enhancements in natESM2
  1. Integration of AI: Leveraging AI, specifically ML, to enhance e.g. data analysis, downscaling, and impact modelling.
  2. Facilitate access to model configurations: Facilitating ensemble simulations over different timescales and ready-to-use configurations for a variety of climate modelling purposes.
  3. Expanded community engagement: Introduction of a summer school, online training programs, and continued workshops and training to foster community growth and knowledge sharing.
  4. Additional RSE positions: Two new RSEs to support advanced software development, data processing, and HPC model integration.
  5. Longer project duration: 48 months to adapt to new computing infrastructures and refine our agile sprint concept together with the growing community. By addressing these critical areas, natESM2 aims to build upon the successes of Phase 1, ensuring that the national Earth system modelling community continues to thrive and innovate in the face of emerging challenges and opportunities. 
Facilitating access to well-defined model configurations  

In Phase 2, we will facilitate access to the following well-defined model configurations to create a truly national ESM system model framework that can be used widely, from education to cutting-edge research and operational use: 

1. Ensemble simulations over 1 to 2 centuries
  • Components: Ocean, atmosphere, and carbon cycle
  • Purpose: Address questions related to climate change, CMIP, IPCC, and decadal predictions 
2. Ensemble simulations over millennia 
  • Components: Ocean, atmosphere, free carbon cycle, ice sheets, and solid Earth
  • Purpose: Paleo-climate modelling and long-term climate futures, teaching
3. (Ensemble) Simulations for air quality 
  • Components: Atmospheric chemistry/aerosol and possibly ocean
  • Purpose: Assess and predict air quality 
4. Kilometer-scale-resolution simulations
  • Components: Atmosphere (≤10 km grid size without cloud and deep-convection parameterizations) and ocean (≤5 km 
    grid size without eddy parameterization)
  • Purpose: Realistic representation of circulation and extremes over years to decades

The project receives fundings from the Federal Ministry of Research, Technology and Space (BMFTR).