Simulation run metadata. Each run metadata component (general attributes, external drivers, internal settings, output settings, simulated time period, and results) has its own metadata. Simulated time period metadata can be applied to different runs. The same internal settings can be applied to different time periods to generate run series. Simulated time period and results are important components of the run metadata for search and discovery.
Space Weather: This paper describes a community-wide effort to understand how “open science” practices can better support research and forecasting in heliophysics and space weather. Open science encourages sharing data, software, and results so that research can progress more quickly and transparently. However, applying these ideas to complex computer models is challenging, because scientific models are much more complicated than typical software and require special care to use and interpret correctly.
To better understand these challenges, we collected input from scientists around the world using an ongoing survey and through in-person discussions at two international meetings held in 2024 and 2025. We also draw on nearly 25 years of experience from the Community Coordinated Modeling Center, which has long supported open access to space science models.
From this work, we identify key obstacles that currently limit how effectively models can be shared, tested, and improved under existing open science guidelines. We organize our findings around four main topics: making models and their results openly usable, improving open and transparent validation of models, supporting open and sustainable model development, and encouraging open collaboration across the community.