2025-B26 Advanced Bioaerosols Network Measurements and Forecasting

Project highlights

CASE studentship with the UK Met Office

Use of novel machine learning techniques to increase the quantity of pollen observations.

Use of novel machine learning techniques to provide bias correction methods which will significantly improve the skill of UK pollen forecasts.

Overview

Bioaerosols are important for human health in indoor and outdoor environments. They are linked to various respiratory illnesses which range in severity from minor to deadly. A high percentage of the UK population has hay fever (allergic rhinitis) due to tree and grass pollen. For many it is an annoyance that can be treated with over-the-counter drugs. However, for a significant percentage of population, the symptoms are far more serious, leading to reductions in work productivity and learning outcomes. Better detection and forecasting of pollen would allow for interventions to be developed that would reduce their risk to human health.

The current methodologies available for the detection of pollen are either expensive and / or time consuming. The UK Met Office currently has only 11 regulatory grade pollen monitoring sites, where pollen is counted daily in May to August and only weekly in March to September. The Met Office has been developing a dispersion-model based capability for forecasting pollen. However, there are still many uncertainties and areas not captured by the model, for example the considerable variability of pollen levels between years. For use in real time forecasting, it is therefore critical to adjust the model according to recent observations, for example by applying bias correction techniques to the model fields. Current pollen observations are limited both in space and in frequency of updates, which significantly impacts the ability to do real time adjustments to the forecast. Low cost, but well distributed monitoring devices with high temporal resolution would allow gaps in the observation data to be filled and subsequently used to improve pollen forecasts.

This PhD combines two rapidly developing technologies. It will bring together distributed internet-of-things (IoT) sensor arrays in combination with artificial intelligence (AI) techniques. Fortunately for this project, pollen has well defined sizes that are distinct to the background aerosol which makes detection possible. Machine learning algorithms will be used to classify the pollen and other species of interest and generate approaches to detect them in real time. This real time detection will allow for improvements in real-time pollen forecasts.

Figure 1: Infographic of proposed methodology for measurement, classification and forecasting of pollen using Internet of Things (IoT) sensors and machine learning.

CENTA Flagship

This is a CENTA Flagship Project

Case funding

This project is suitable for CASE funding

Host

University of Birmingham

Theme

Climate and Environmental Sustainability

Supervisors

Project investigator

Francis Pope (University of Birmingham, [email protected])

Co-investigators

Lucy Neal (UK Met Office, [email protected])

How to apply

Each host has a slightly different application process.
Find out how to apply for this studentship.
All applications must include the CENTA application form.
Choose your application route

Methodology

Low-cost sensors will be deployed at a range of field locations in a limited area of the UK. Various statistical and machine learning methods will be used to analyse data collected from these instruments. Results from the sensors will also be compared to regulatory grade pollen traps to ensure robustness of results.

Methods which allow incorporation of the sensor data into pollen forecast fields to improve the forecasts will be investigated. This will include considering existing bias correction algorithms or newer machine learning techniques. Network optimisation methods will also be used to determine recommendations for the number and location of sensors that would be needed to provide improved real-time pollen forecasts for the entire UK.

Training and skills

DRs will be awarded CENTA Training Credits (CTCs) for participation in CENTA-provided and ‘free choice’ external training. One CTC can be earned per 3 hours training, and DRs must accrue 100 CTCs across the three and a half years of their PhD.

Full training will be provided on the required modelling packages and programming languages. The candidate will become an Integral member of the Birmingham Institute of Forest Research (BIFoR) and the Environmental Health Sciences theme of the University of Birmingham, benefitting from the lively intellectual atmosphere of both communities. The candidate will be encouraged to attend cognate MSc modules in atmospheric science and lectures in plant ecophysiology at the University of Birmingham as their training and development needs require. The student will benefit enormously from collaboration with industrial and governance colleagues. The student will visit the Met Office in Exeter for extended periods and will benefit from working within the atmospheric dispersion and air quality team, to learn about real world use of the pollen forecast model, the existing bias correction and verification methods, as well as to understand and impact upon the UK’s pollen network works.

Partners and collaboration

The student will benefit from extensive links between the project with governance partners, including the UK Met Office (CASE partner), UK Health Security Agency (UKHSA) and the Environment Agency. The Pope group has extensive links with the wider European aerobiology community through EU grants, such as the ADOPT EU Cost action.

Further details

Prof. Francis Pope ([email protected]) will be delighted to take informal questions about the project. Website – www.francispope.com.

 To apply to this project: 

You must include a CENTA studentship application form, downloadable from: CENTA Studentship Application Form 2025.

You must include a CV with the names of at least two referees (preferably three) who can comment on your academic abilities.

Please submit your application and complete the host institution application process via: https://sits.bham.ac.uk/lpages/LES068.htm.   Please select the PhD Geography and Environmental Science (CENTA) 2025/26 Apply Now button. The CENTA Studentship Application Form 2025 and CV can be uploaded to the Application Information section of the online form.  Please quote CENTA 2025-B26 when completing the application form.

 Applications must be submitted by 23:59 GMT on Wednesday 8th January 2025.  

Possible timeline

Year 1

Literature survey and review paper. Exploration and development of different pollen detection techniques, including machine learning approaches, and low cost air particulate matter sensors. Development of initial method to combine low cost pollen sensor data with model fields. Pilot UK based measurement campaign. Attend national UK conference.

Year 2

Further development of pollen detection techniques to allow for the input of multiple sensor data, including the use of machine learning techniques to extract pollen information from sensor data. Extensive UK based field campaign during both the tree and grass pollen seasons. Combining pollen data from regulatory sites and sensor measurements to create hybrid pollen maps.

Year 3

Data analysis and interpretation. Implementation of field data into model. Discussions and analysis will be finalized here, and final outputs will be published. Presentation of results at international conference e.g. AGU in San Francisco, USA. Thesis preparation and viva.