Project Description
UK river water quality is a pressing political, public and environmental concern: only 15% of English rivers meet ‘good’ status under the Water Framework Directive, and water companies and the government face mounting pressure to improve waterways. The picture is similarly stark globally, where freshwater vertebrate populations have declined by 85%, driven by pollution, habitat loss, agricultural intensification and climate change. Freshwater science urgently needs new tools to diagnose declining water quality, attribute it to specific stressors (e.g. sewage pollution, abstraction, diffuse agricultural runoff) and inform effective mitigation. Monitoring is transitioning from ‘grab’ sampling at weekly to monthly intervals to continuous, sensor-based monitoring at sub-hourly resolution. This shift, exemplified by the UK’s Section 82 (Environment Act 2021) programme to deploy 20,000 continuous sensors nationally, generates high-volume, multivariate datasets that reveal short-lived pollution events and natural diurnal cycles invisible to grab sampling. However, existing manual interpretation methods cannot keep pace with this data volume and complexity, creating a critical bottleneck between data collection and actionable insight.
This project will develop novel data-scientific methods to interpret multivariate, continuous water quality datasets at national scale. The student will: (1) build data-reduction and dimensionality-reduction approaches that enable large numbers of sensors to be compared systematically, identifying recurring ‘typologies’ of water quality response; (2) adapt event-detection methods from bioacoustics and seismology to automatically identify discrete events, such as sewage spills or storm-driven pollution pulses, within continuous sensor records using machine learning; and (3) apply these methods to a UK-wide compilation of sensor datasets, assembled by the student, to characterise the land-use, climatic and hydrological drivers of water quality nationally for the first time. The project directly addresses NERC’s priorities in environmental data science and freshwater health, and translating research into environmental management practice, responding to an urgent need created by the UK’s rapid expansion of continuous monitoring infrastructure.
Working closely with CASE partner The Rivers Trust, the student will co-develop a practitioner toolkit that translates these methods into a standardised workflow for extracting insight from continuous monitoring data. Needs identified by the CASE supervisor will steer project development, ensuring the research responds directly to practitioner requirements. The resulting toolkit will be deployed across The Rivers Trust’s network of catchment managers, providing a direct route from methodological innovation to real-world impact. During their 6 month placement with The Rivers Trust, the student will bring their expertise in continuous monitoring data science to the organisation’s data science team, contributing to an active project with Defra and the water industry on establishing ‘baseline’ water quality conditions against which pollution events can be identified, a key question for optimising the future deployment of continuous sensors nationally. Specialist expertise in continuous monitoring and data science is essential to this work, and the student will be uniquely placed to fill this gap, gaining direct experience of how research translates into environmental policy and practice.
Project Specific Training
he student will join the active Surface Processes research cluster within the Department of Earth Sciences UCL. This research group (containing both the main and co- supervisors) meets weekly with paper discussions and presentations on relevant research topics spanning hydrology, geomorphology and chemical weathering. These group meetings provide a supportive environment for students to become familiar presenting their work and critiquing papers. This specific research group is also complemented by other relevant research groups within the department for instance the biweekly LOGIC geochemistry research seminars and the GoGo geophysics seminars.
The student will be actively encouraged to attend a diversity of relevant conferences and meetings firmly establishing them within an academic environment, furthering an academic career should this be their goal. They will regularly attend the annual meeting of the British Hydrological Society, will attend the Annual Meeting of the European Geosciences Union in Vienna at least once, as well as smaller local conferences such as the London Freshwater Group.
The student will be integrated into the S82 working groups, which bring together industry partners and technical experts coordinating national implementation of the sensor network. This will develop their skills in cross-sector collaboration, communicating with non-academic stakeholders, and understanding the practical and commercial realities of environmental monitoring at scale, alongside core data science skills.
Potential Career Trajectory
This primarily computational project will deliver strong transferable computational skills relevant for non-academic career pathways. For instance the student will become familiar with best practice in scientific software development including advanced Python programming, version control & collaborative software (GitHub), and high performance computing. These directly relevant skills complement the wider transferable skills from any PhD programme including scientific writing, presentation skills, and independent research. The applied nature of the programme, quantitative skillset and industrial placement position the finishing student well to graduate into a non-academic applied role in environmental management should they choose this route.
Project supervisor/s
Alex Lipp
Earth Sciences
University College London
a.lipp@ucl.ac.uk
https://profiles.ucl.ac.uk/99868-alexander-lipp
Matthew Fox
Earth Sciences
University College London
m.fox@ucl.ac.uk
https://www.ucl.ac.uk/mathematical-physical-sciences/earth-sciences/people/dr-m…
