Apply now

Find out more about the different routes to entry and our eligibility criteria

Historical phenological baselines for UK birds

image
Project Description

The timing of egg laying in wild birds has advanced over the last 40-50 years in response to climate change. The absence of older data means we lack the deeper time perspective we need to contextualise the rate and magnitude of these contemporary changes. This is critical for a more complete understanding of the impacts of climate change. A deeper time perspective is possible using data from historical egg collections, but to date this has been prohibitively difficult because it may take several years to extract the data from physical records. These constraints are being transformed by artificial intelligence (AI); it is now feasible to extract data from thousands of physical records in just a few months. Our project aims to further develop and exploit these approaches to produce historical baselines on the timing of egg laying for UK wild birds for the first time. We will: (1) further develop and refine AI approaches to extract data from ~100k record cards relating to ~200 plus UK species from the egg collection held by the Natural history Museum in Tring, (2) use the resulting data to produce historical baselines on the timing of egg laying in UK birds covering a window of about 200 years, and (3) compare these baselines with contemporary data to estimate the rate and magnitude of ecological change and understand the drivers of variation between species. We have recently developed a pipeline for extracting data from egg records cards based on imaging and AI approaches.

Research themes
Project Specific Training

We will undertake a training needs analysis with the student soon after they start. This will help us understand their basic skills and any gaps that need to be addressed. We know the student will require skills development in three key areas: (1) working with the egg collection at Tring, (2) AI/CV approaches and their application to extracting data from collections’ records, and (3) the management and analysis of ecological data. Our supervisory team covers all these major training areas – collections, AI/CV and data management/ analyses. Primary supervision will be provided through regular meetings, etc with the student, and they will manage more specialised supervisory input from the wider team as required. We have a workplan that sets out these elements and will ensure the student finishes on time.

At this stage, we see student development being primarily delivered through the supervisory group and the wider teams of which they are part. This will provide the student with the disciplinary skills needed for the various components of the planned research, but also access to a rich community of researchers to learn from in a range of other ways e.g. communication, paper writing, etc. We plan to explore opportunities in more detail after the student starts as part of the assessment of their training needs. We recognise that the student might benefit from access to more formal PGR training, but this will need to be tailored to the needs of the student. For example, if the student has a background in computer/data science, courses in ecology might be helpful; or vice versa if they have a background in the biological sciences. Again, we will assess these needs after the student starts.

Potential Career Trajectory

We see three major potential pathways:
1. Academia - the role of AI in environmental science is only going to grow, so the student would be well placed to develop a research career that integrated AI approaches into any 'big data' problem related to the environment.
2. Natural History Museums - the role of AI is also going to grow significantly within the Museum/collections sector, so the student would be well placed to develop a career in collections development, management or research.
3. Data science - the student will develop strong, generic data science skills during the PhD that would be potentially relevant to any sector for which 'big data' management or analysis is important. 

Project supervisor/s
Professor Ken Norris
Science Group
NHM
k.norris@nhm.ac.uk
Tim Newbold
Genetics, Evolution & Environment
UCL
t.newbold@ucl.ac.uk
Supervision balance
70:30