Professor Peter AtkinsonFaculty Dean
Peter’s research can be characterized as "spatial data science” and it arises out of the application of quantitative geography to remotely sensed data. Specifically, it involves the application of space-time statistics and geostatistics, machine learning and numerical modeling, to Earth observation (EO) and other environmental and social data, to answer a wide range of science and social science questions.
These science questions mostly relate to understanding:
- Spatial and space-time sampling effects
- Disease transmission systems, especially vector-borne systems
- Global vegetation and land cover changes
- Natural hazard impacts and risks
Peter’s research is, thus, highly inter-disciplinary and spans a wide range of statistical and numerical techniques and a wide range of science and social science domains.
The four most significant themes of Peter’s research currently are:
- The development of novel image “downscaling” and image fusion techniques for application to time-series of remotely sensed images, based on explicit models of the spatial (and space-time) sampling framework.
- Remote sensing of changes in global vegetation phenology (the growing rhythm of plants, e.g., the start of spring, end of autumn) to increase our understanding of how these are driven by climate changes. Related to this theme, the use of time-series data on vegetation productivity to forecast crop yield (and yield gaps) in support of reducing food insecurity.
- Spatial epidemiology of vector-borne disease transmission systems, such as malaria and Trypanosomiasis, based on: (i) agent-based models of the transmission dynamics such as to connect ecosystem changes with disease and poverty outcomes, and (ii) Bayesian mixed regression models to predict risk in space-time so that interventions can be targeted.
- Spatial modeling of physical natural hazards and their impacts, including: flood forecasting based on Kalman-filter variants; landslide susceptibility mapping based on mixed regression models; and, together with colleagues at Southampton, near-Earth object impact and risk simulation based on Newtonian orbital dynamics.
Peter is keen to supervise exceptional PhD students with a strong mathematical, statistical or computer science background, who have an interest in these research topics. If you are interested in applying for a PhD in any of these topics please contact Peter at firstname.lastname@example.org.
Peter has published over 200 peer-reviewed international scientific journal articles on these topics, and authored or edited nine books. He has also published around 50 refereed book chapters, and edited nine journal special issues. He has chaired or co-chaired several major international conferences including GeoComputation in 2003, GeoENV in 2008 and RSPSoc in 2015. He has led multiple large grants and supervised over 50 PhD students. His Thompson H-index is 52 in Google Scholar and 36 in WoS.
Peter is Dean of the Faculty of Science and Technology at Lancaster University. He was previously Associate Director of REF Strategy (the academic lead) in the run up to REF2014 at the University of Southampton and before that he was Head of the School of Geography in Southampton for five years.
Peter is currently Visiting Research Professor at Queen’s University Belfast and Visiting Professor at the University of Southampton. Peter was the holder of the Belle van Zuylen Chair at Utrecht University (2015-16), and Visiting Fellow at Green-Templeton College, Oxford University (2012-14).
He is the 2016 recipient of the Peter Burrough Medal of the International Spatial Accuracy Research Association (ISARA).
Peter is Associate Editor of Computers and Geosciences and sits on the editorial boards of a raft of journals including Geographical Analysis, Spatial Statistics, the International Journal of Applied Earth Observation and Geoinformation, and Environmental Informatics.
Explainable AI for UK agricultural land use decision-making
01/12/2019 → 30/11/2020
IAA: Deep Learning in Massive Area, Multi-scale Resolution Remotely Sensed Imagery EPSRC Code
02/09/2019 → 30/11/2020
Newton Fund PhD Placement - Xuyu Bai
01/04/2019 → 31/03/2020
Advancing Drought Monitoring, Prediction, and Management Capabilities (Workshop)
17/09/2018 → 20/09/2018
DSI: Data Science of the Natural Environment
16/04/2018 → 15/04/2024
Proximity to Discover: Industry Engagement for Impact
31/03/2018 → 30/11/2019
NERC (External organisation)
Membership of committee
SCC (Data Science)
Geospatial Data Science
- DSI - Environment
- Geospatial Data Science
- Lancaster Intelligent, Robotic and Autonomous Systems Centre