Establishing Community Requirements of Hierarchical System Development for Earth System Models

The concept of Hierarchical System Development (HSD) provides an efficient pathway for model development and scientific advancement. It provides a framework to the research community with multiple entry points for research efforts spanning simple to complex models. These entry points may exist on different axes, defined by i.) sample size, ranging from a single case study to a large number of runs, ii.) spatial scale, from coarse to fine grid, global to regional, or nested or variable resolution domains, iii.) simulation realism, from idealized simulations to fully-coupled models, and iv.) simplification of models through mechanism or interaction denial, such as substituting a component or mechanism with prescribed data. Ideally, a researcher can use the system from available entry points based on their scientific needs and purpose. Ultimately, HSD connects our understanding from idealized models to comprehensive models, and improves understanding of model biases through the ability to simplify the system to select key processes or interactions. It also enables research to operations (R2O) by providing testing pathways for innovations and updates from the research community to developers. The Developmental Testbed Center (DTC), in collaboration with the Earth Prediction Innovation Center (EPIC), prepared a white paper describing the long-term vision of HSD for the UFS and the plan for its phased implementation. The paper outlines the proposed hierarchical axes relevant to the UFS, HSD capabilities that currently exist in the UFS, and recommendations for future development by EPIC aligned with each axis. A supplemental survey on HSD for UFS was distributed to a broad research and operational community to collect valuable insight and feedback on participants' experience with, and use of, HSD tools, testing infrastructure, current and future capability needs, and perceived gaps to help inform the future HSD within the UFS. Themes that emerged from the survey include the need for HSD development to be included in all model components, development of highly configurable workflows for testing with detailed documentation, the ability to ingest a variety of input data for initial and boundary conditions, supporting nesting and/or variable resolution capabilities, and accessibility and ease of use to the community. The results were used to inform and compose the white paper. While HSD exists to some extent community-wide, closer collaboration across the research and development communities will accelerate the progress towards improved model performance.