EvaSuite – Evaluation Suite for Climate Data

• Documentation
• Installation
• Configuration
• Running EvaSuite
• Examples
• FAQ
• EvaSuite Test Data
• EvaSuite Extensions

Introduction

Evaluation Suite (EvaSuite) is a modular software framework for the standardised evaluation of regional climate model (RCM) simulations against observational and reference datasets, including model sensitivity analysis. It is designed for high-performance computing platforms and addresses the need for transparent, reproducible, and flexible evaluation workflows for RCMs. EvaSuite combines data harmonisation, spatial and temporal remapping, statistical evaluation, and visualisation within a unified fully configurable, end-to-end workflow. EvaSuite is maintained by the CLM-Community and is designed for reproducible use on HPC systems as well as on local workstations (under developing).

All evaluation workflows are controlled through YAML configuration files defining:

• system settings and external dependencies;
• model and reference datasets;
• preprocessing options;
• evaluation metrics;
• visualisation settings.

For the complete user documentation, see the EvaSuite documentation.

Key Features

• End-to-end evaluation workflow
  • Data harmonisation, remapping, metric computation and visualisation.
• YAML-driven configuration
  • Reproducible and automation-friendly workflows.
• HPC-ready design
  • Supports direct execution, interactive use and SLURM batch submission.
• Local execution
  • Simple examples can also be run on a local workstation with local dataset catalogues.
• Flexible data support
  • CF-compliant NetCDF datasets from models, reanalyses and observations.
• Advanced diagnostics
  • Bias, RMSE, correlation, quantile-based diagnostics, significance testing and other evaluation metrics.
• Integrated visualisation
  • Maps, time series, QQ plots and Taylor diagrams.
• Modular extensions
  • Additional tools and workflows are maintained in separate EvaSuite extension projects.

General Information

The history and recent updates of EvaSuite provide an overview of the framework evolution and newly introduced features. Detailed instructions for installation and system setup are described in the Installation Guide, including environment preparation and known issues. EvaSuite can be executed in different modes (main, interactive, debug), as explained in the running documentation, allowing flexible usage for both batch HPC workflows and exploratory analysis. User-specific customization of plots, maps, and diagnostics is controlled via configuration files, documented in the user settings sections for plots, maps and Taylor diagrams. A comprehensive collection of reproducible workflows is available in the examples, demonstrating typical evaluation setups and wrapper-based pipelines. Additional technical details, module descriptions, and theoretical background are provided in the full documentation. For advanced applications and extensions (e.g. SPICE workflows), see the dedicated SPICE documentation. Modular extensions presented in separate projects GeoPackage, GriddedStations, ScoPi and UrbanPlot

Getting help

Before opening an issue, check:

• the documentation;
• the FAQ.

If the problem remains unresolved, open an issue on GitLab.

Citation

If you use EvaSuite in your research, please cite:

Petrik, R., Geyer, B., Churiulin, E., Rockel, B., & Braun, C. (2026). EvaSuite (Evaluation Suite for climate data) (v1.0.1). Zenodo. https://doi.org/10.5281/zenodo.19223685

BibTeX

@software{petrik2026evasuite,
  author       = {Petrik, Ronny and Geyer, Beate and Churiulin, Evgenii and Rockel, Burkhardt and Braun, Christoph},
  title        = {EvaSuite (Evaluation Suite for climate data)},
  year         = {2026},
  version      = {v1.0.1},
  publisher    = {Zenodo},
  doi          = {10.5281/zenodo.19223685},
  url          = {https://doi.org/10.5281/zenodo.19223685}
}

Contributing

Contributions of new diagnostics, workflows and features are welcome. Please follow the CLM-Community development and contribution conventions and coordinate larger changes through the project issue tracker.