Live demonstration of the UK-EMS

Jonathon Cooper, Michael Holloway (UKCEH)

A live demonstration of the UK-EMS website was provided.  It showed how the data from the National Atmospheric Emissions Inventory has been made accessible.  It covered:

Discovery: searching for emission and activity data, point source data and notebooks, and viewing metadata
Visualisation: interactive map and timeseries chart for pollutant emissions, with point source overlay and snap sector filtering.
Download: netcdf download from the visualisation page with spatial, temporal and snap sector sub-setting 
API: data is made available via an API.  Its documentation is available on the website, and dynamic URLs that help start using it are made readily available throughout the website.  Python code and sample notebooks were demonstrated.
Datalabs: this is an on-line analytic environment where code can be developed principally through Jupyter notebooks.  It has fast access to DUKEMS data.  An introductory notebook, available in Datalabs and Github, was used to show how to import data from the API and view it ready for integrating into your own code.


Scientific development & improved emission calculation methods: Road transport

David Carslaw (University of York), Gregor Stewart (Imperial College London)

New speed-emissions curves for NOx have been derived from application of remote sensing data to vehicle drive cycles.  The new curves have largely eliminated bias in our urban air quality model in London, when the modelled outputs are compared with monitoring site data in 2019.  The largest emissions changes seen are for heavy vehicles (HGVs and buses), where Euro VI COPERT functions appear to substantially underestimate NOx emissions. 


Scientific development & improved emission calculation methods: Improvements and Additional Information for UK-EMS from the National Atmospheric Emissions Inventory

Matthieu Pommier, Tim Murrells, Ioannis Tsagatakis et al. (Ricardo)

This presentation provides an overview of the data that underpins the EMS tool. The work is based on the expertise used to compile and maintain the National Atmospheric Emissions Inventory and the Greenhouse Gas Inventory Programmes for UK Government.

This work includes the timeseries in emissions data from 2005 to 2030 for stationary and mobile sources covering a range of pollutants and the presentation describes two examples where more detailed data are provided on shipping and on rail emissions. The presentation also explains the sources which were considered and not included in the NAEI, such as the cooking organic aerosols and road dust resuspension of PM emissions. Additional data to standard NAEI data are also described, including stack parameters for point sources, the development of a detailed NMVOC chemical speciation, the mapping of this speciation to compounds described by different chemical schemes used in chemical transport models and the improvement of the PM chemical speciation. These speciations are developed for the years from 2005 to 2019 at GNRF, NFR and SNAP levels and available through excel tables.


Scientific development & improved emission calculation methods: Modelling emissions time series

Sam Tomlinson (UKCEH)

It is important to model emissions not only in space but in time as well, to understand pollution events and to best mitigate for detrimental effects from air pollution. In DUKEMS we have modelled sub-annual temporal resolutions of emissions using a novel method with a statistical approach. Generalised Additive Models were used to smooth various sources of activity data into diurnal, day of week and monthly temporal profiles which can be either be applied to emissions and/or integrated directly into atmospheric chemistry transport models. This is a dataset made in the UK context in a framework that can incorporate many sources of input data and output temporal profiles for any pollutant


Uncertainty assessment in DUKEMS and beyond

Richard Claxton (Aether), Tom Gardiner (NPL)

Richard Claxton (Aether) and Tom Gardiner (NPL) presented information on how uncertainty and data quality have been handled within the development of the UK-EMS. Quantifying uncertainties for emissions data is challenging for a number of reasons – partly due to the different mathematical approaches available, but also because of the scope of the emissions data (e.g. how to best represent temporal and geographical/mapped uncertainties). Importantly, uncertainty estimates that are submitted as part of the UK NAEI are produced under specific reporting purposes, and to give an indication of inventory accuracy over time. The uncertainty data is not typically appropriate for direct use in determining subsequent sensitivity of emissions models. As such, the project team has focused on developing a flexible data model within the UK-EMS so that uncertainty metrics can be brought into the system by the user. The system also provides detailed information and metadata fields to users related to data origins and provenance so that individual data sets, or data points are traceable. This was a key requirement highlighted by the model user group during system build.

Future developments and governance

Richard Claxton (Aether)

An interactive session was held to gather feedback from the audience and users on potential future developments and governance options for the UK-EMS. The audience was overwhelmingly positive about the benefit UK-EMS will provide. However, it was highlighted that uptake of the system will take time as modellers will need to fully test and understand the system if it is to replace or add to their existing suite of options. As such, some form of moderation and governance system is considered critical if the UK-EMS is to become a national staple for ongoing access and interrogation of UK emissions data.