Model Evaluation Toolkit

Free toolkit for evaluating atmospheric dispersion models

The Model Evaluation Toolkit is a free toolkit for evaluating local and regional atmospheric dispersion models.

Key features of the Model Evaluation Toolkit:

• Compare one or more modelled datasets with in situ observed data
• Analysis can be based on pollutant concentrations and/or forecast indices and alerts
• Output includes statistics and attractive report-ready graphs
• Group analysis results by monitoring station, station type, by pollutant or by model
• Wide range of supported modelled data formats including ADMS, CMAQ, CAMx, WRF-Chem, EMEP, CHIMERE and CAMS
• Easy to use graphical user interface
• Uses openair tools
• Graphical output includes state of the art target plot as developed by the FAIRMODE initiative
• Observed data from UK monitoring networks can be downloaded and imported automatically

Click here to download the Model Evaluation Toolkit (version 5.4, released June 2026):Model Evaluation Toolkit (.zip, 80 MB)

Once downloaded, you should extract the contents of the zip file to your computer and follow these installation instructions: (.pdf, <6 MB)

You will find a comprehensive User Guide here:(.pdf, 5 MB)

Model Evaluation Toolkit Version 5.4 Release Notes

Version 5.4 is a major new version of the Model Evaluation Toolkit, and introduces a wide range of enhancements that improve flexibility, data compatibility and analytical power.

The update significantly expands supported modelled concentration data formats. The tool is now compatible with a range of commonly used gridded datasets including MAQS, WRF-Chem, CMAQ, CAMx, CHIMERE and EMEP, alongside Defra background map data.

Data handling has been overhauled, with improved integration of monitoring networks using openair tools. Observational air quality data can now be automatically downloaded across multiple networks, while new metadata tools help users select appropriate monitoring sites more efficiently. Meteorological data inputs have also been added, enabling advanced evaluation analysis visualisations such as polar plots.

The Model Evaluation Tool can now analyse ensemble median datasets and composite pollutants such as the ratio of NO2 to NOX, or total oxidant (NO2+O3). New visualisation options – like statistical scatter plots and openair time variation plots – provide deeper insights into model performance. Additional metrics, including mean square error components, strengthen model evaluation techniques. Air quality forecast analysis capabilities have been enhanced, with new tools to identify pollution episodes based on user-defined thresholds, and compatibility with multiple air pollution index scales such as the UK’s Daily Air Quality Index (DAQI) and the European Air Quality Index.

The Model Diagnostics Tool has greater flexibility, including selectable datasets, date ranges, and new polar plot functionality. Improvements to plotting controls and data markers offer more detailed visual analysis.

A universal shift to more flexible time-averaging and updated R package management completes the upgrade.

For more information, please see the Model Evaluation Toolkit version 5.4 User Guide.