20 May 2026Atmospheric dispersion model evaluation methods report published

A key part of atmospheric dispersion model usage is evaluation using air concentration measurements. A wide variety of statistical and graphical evaluation methods are available, with no single method able to capture all aspects of model performance. Model users can find it difficult to decide on the most suitable methods for evaluating their own models, including how to define "acceptable" model performance. Policy makers also need guidance on how to interpret the results of atmospheric dispersion model evaluation.

CERC have recently completed the study 'A review of methods used to assess the performance of atmospheric dispersion models', commissioned by the UK Atmospheric Dispersion Modelling Liaison Committee (ADMLC). The report provides a comprehensive review, covering evaluation methods and their suitability for different modelling applications. The study analysed 70 evaluation metrics across nine case studies and more than 30 models. Statistical metric thresholds for "acceptable" model performance have been proposed for non dimensional metrics, with more and less stringent criteria suggested where possible. Proposed thresholds were also compared with established acceptance criteria. CERC's freely available Model Evaluation Toolkit was used to perform the statistical calculations, and the updated version of this tool will be available in due course.

The full report can be accessed here.

A related in-person workshop has been organised for the 11^th June 2026, with the aim of gathering feedback from the atmospheric dispersion modelling community on the suitability of the threshold criteria proposed in the study. The workshop agenda is available here. Sign up by completing this form: https://forms.cloud.microsoft/e/6MRPrW3Yzg

The image shows an example of calculating thresholds for Normalised Absolute Error (NAE): (a) interquartile range results of NAE for the evaluated case studies, (b) these results presented in tabulated form, and (c) case studies assessed against acceptance criteria from Chang & Hanna (2004, 2012).

11 May 2026Workshop on Atmospheric Dispersion Model Evaluation Methods and Tools

The Atmospheric Dispersion Modelling Liaison Committee (ADMLC), in collaboration with CERC, are hosting a workshop focused on atmospheric dispersion model evaluation metrics, including new proposals for model performance threshold criteria. Further information, including the workshop agenda, is given here.

The workshop builds on the outcomes of a recent ADMLC-commissioned study by CERC, which reviewed a wide range of statistical and graphical methods used in atmospheric dispersion model evaluation. The study analysed 70 evaluation metrics across nine case studies and more than 30 models, covering applications such as regulatory assessment, compliance, accidental release and air quality forecasting.

A key outcome of the project is proposed performance threshold criteria for a large number of non-dimensional model evaluation metrics. Where possible, the study compares thresholds to criteria already used within the atmospheric dispersion modelling community e.g. from Chang and Hanna (2004, 2012). This workshop will provide an opportunity to:

• Present proposed threshold criteria and associated derivation methodologies
• Discuss threshold suitability and applicability across different modelling contexts
• Gather feedback to support potential adoption within the atmospheric dispersion modelling community

CERC will also showcase the Model Evaluation Toolkit, which was used to perform the case study model evaluation calculations. The Toolkit calculates the wide range of statistical metrics presented in the report and generates supporting graphics. It is available as a free resource to support consistent model evaluation practices. A new version will be released before the workshop.

This workshop is for practitioners, researchers and regulators involved in atmospheric dispersion model evaluation who are interested in improving the consistency and robustness of their techniques. There is no charge for the event but registration is required.

• Thursday 11 June 2026, 10:00‐16:00
• Harwell, Oxfordshire
• Please register by completing this form: https://forms.cloud.microsoft/e/6MRPrW3Yzg
• Deadline: Thursday 21 May 2026

The image shows examples of the types of figures and tables presented in the report: a decomposed error metric; metric results tabulated for all case studies; a boxplot for calculating proposed thresholds.

29 Apr 2026Tribute to Julian Hunt

It is with great sadness we announce that Julian Hunt died on 20 April 2026.

As a brilliant scientist, Julian applied his mind and inexhaustible energy to a wide range of fluid mechanical problems. These ranged from fundamental studies of turbulence near surfaces and around obstacles, including his seminal papers on the rapid distortion theory of turbulence, to many ground-breaking studies with more immediate applications, including: turbulent airflow over hills; structure of the atmospheric boundary layer; atmospheric dispersion of pollutants over complex terrain; and airflow in urban and coastal environments.

Julian was always keen on communicating his research and applying it to practical applications. This led to him being a founding director of Cambridge Environmental Research Consultants in 1985; as he told me, he was motivated by his research being applied in the USA, "so why not in the UK?" At CERC, he was the leading force in the development of a range of fast environmental models incorporating state-of-the-art science. The first of these was FLOWSTAR for turbulent flow over hills, accounting for stratification and roughness. However, the most well-known is, of course, the Atmospheric Dispersion Modelling System, ADMS. ADMS was first released in 1991 and is now widely used across the world. Making extensive use of his theoretical advances, Julian constructed a model that was pioneering in its treatment of the atmospheric boundary structure and in extending treatment of Gaussian plume dispersion modelling to complex flows. Its advanced treatment of dispersion in inhomogeneous flow and turbulence over complex terrain has still not been replicated in any other Gaussian plume dispersion model.

Julian remained as Chairman of CERC until as recently as 2022, when he was still full of new ideas! He will be greatly missed.

David Carruthers, 27 April 2026

Image shows Julian (right) outside CERC offices in Cambridge in the 1990s, alongside other CERC Directors David Carruthers, Paul Linden and Tim Newton.

24 Apr 2026Air Quality Lifecourse Assessment Tool (AQ-LAT) featured in Air Quality News

The Air Quality Lifecourse Assessment Tool (AQ-LAT) has been featured in the April 2026 issue of Air Quality News. The article outlines how the tool links air quality modelling outputs with estimates of health and economic impacts to support policy development.

Developed by the University of Birmingham, the tool combines high-resolution air pollution data with population datasets and established epidemiological relationships to quantify outcomes such as premature mortality, disease incidence, healthcare costs, and productivity losses.

In current applications, high-resolution air pollution maps are generated by CERC using ADMS-Urban. These outputs provide the exposure data required for AQ-LAT to calculate health and economic impacts at a local scale.

This method has been applied to evaluate policy scenarios in the West Midlands, as a part of the WM-Air project, and in Oxfordshire, by OxonAir (a partnership between Oxfordshire County Council and the county's city and district councils).

This approach allows users to quantify the impacts of air quality interventions in terms of both pollutant concentrations and their associated health and economic outcomes. If you would like to know more about combining ADMS-Urban with AQ-LAT, please get in touch.

14 Apr 2026HARMO conference papers & other recent co-author publications

The outcomes of several projects have recently been shared through international conference presentations, co-authored publications and technical reports. We were co-authors on four peer-reviewed papers during 2025, with another article already published in 2026. These studies reflect ongoing collaboration with universities, government and research institutes across a range of applied air quality topics.

CERC's Dr Martin Seaton and Dr Victoria Hamilton attended the HARMO conference Harmonisation within Atmospheric Dispersion Modelling for Regulatory Purposes), which aims to improve consistency and best practice in atmospheric dispersion modelling.

Martin presented work assessing the impact of thermally driven flows on regulatory air dispersion modelling (paper here). This is particularly important in complex terrain and coastal environments, and forms part of CERC's contribution to the Environment Agency's research on the thermal transport of air pollution from regulated industries.

RIVM (the Dutch National Institute for Public Health and the Environment) presented results from a study comparing eight widely used atmospheric dispersion and deposition models, including ADMS, to which CERC contributed. The study investigated ammonia and nitrogen oxides at the local scale (paper here) and highlighted the variability and complexity associated with modelling deposition.

Victoria presented an extension of the study, evaluating alternative concentration-dependent deposition (CDD) methods against current UK guidance (paper here). The image shows the modelled concentration and deposition results for the four CDD methods tested. Depending on model choice, resulting deposition can differ by up to a factor of 10 for low concentration values, and a factor of 3 for high values. Models accounting for bi-directional exchange of ammonia ('Big-leaf' and DEPAC) predict more accurate values for this case study.

The presentations and extended abstracts are now available on the HARMO website:

https://www.harmo.org/conference.php?id=23

Further recent co-authored work includes:

• Ongoing research with Dr Zhong (University of Greenwich) in the West Midlands, including studies of urban heat islands and scenario modelling of bus electrification as a part of the WM-Adapt, WM-Net-Zero and WM-Air projects
• Work led by Dr Tom Clemens (University of Edinburgh) examining the effects of air pollution on birth outcomes in Scotland

The studies make use of a range of CERC tools, including ADMS-Urban, ADMS-Urban Temperature & Humidity, ADMS 6.

Links to all the publications are available here:

https://cerc.co.uk/environmental-software/CERC-coauthor-publications.html

We look forward to continuing these collaborations during 2026.