Community modeling: supporting large-domain water resources modeling through an open-source model configuration toolbox
Abstract
Configuring Earth System models can be a cumbersome task, especially for larger domains. It is not uncommon to hear claims that model setup takes 80% of time, leaving only a meager 20% for scientific analysis. Despite internal differences between numerous existing models, many model configuration tasks are very similar. The modeling community would benefit from standardizing these tasks wherever possible and openly sharing the necessary model configuration code. This will improve the transparency of modelling results and potentially lead to large efficiency gains for individuals, groups and the community as a whole. We have created an open-source, reproducible workflow for local to global model configurations of the hydrologic Structure for Unifying Multiple Modeling Alternatives (SUMMA) and the river routing model mizuRoute. The workflow uses fully open source input data, model code and configuration code to simulate the terrestrial water and energy balance. Based on our experiences with creating this workflow, we define the possible structure of an open-source model configuration toolbox that can be used to harmonize model configuration code for different models and datasets. We first distinguish between model-agnostic and model-specific tasks, and further separate the model-agnostic tasks into dataset-specific and dataset-agnostic tasks. Separating model configuration tasks in such a manner allows efficient re-use of existing code and inclusion of new datasets and models with limited effort. We view our workflow as the first step in developing such a community-driven modular model configuration framework, in which new model setups can be generated with a fraction of the effort currently required, and that increases the transparency and reproducibility of hydrologic modeling studies.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2021
- Bibcode:
- 2021AGUFM.H43D..05K