Three speakers from clinical, environmental and geographical domains speaking about a diverse range of methods they use to simulate their phenomena of interest.
Ann-Kristin Koehler is a Research Fellow in the Leeds School of Earth and Environment who has worked on crop modelling.
Verity Tether is a Doctoral researcher in the Leeds School of Geography and has used agent-based modelling to investigate crime patterns.
Tom Lawton is a Consultant in Critical Care & Anaesthesia and Head of Clinical Artificial Intelligence at Bradford Royal Infirmary who has modelled patient flows using discrete event simulation.
This special session will also include our Annual General Meeting, wherein we will report on the year's activities and elect members of the committee for the coming year.
We will have an officer vacancy to fill on next year's committee. The officer positions are Chair (currently, John Bates), Secretary (currently, Ciarán McInerney), and Treasurer (currently, Paul Baxter). Current nominations are below but we strongly encourage nominations that will provide novel perspectives on the committee:
- Chair: John Bates
- Secretary: Ciarán McInerney
- Treasurer: none
Applicants should provide a short statement, as it will help the committee to understand who is applying, why they are interested in joining, any prior experience they have and what they will bring to the role. A paragraph (up to 300 words), summarising their background and why they are interested would be acceptable.
Send statements to
rssleedsbradfordgroup@gmail.com
We also have some vacancies for ordinary committee members, who have a voice on the committee without officer roles. Similar to the officer position, applicants should provide a short statement, as detailed above.
Three speakers from clinical, environmental and geographical domains speaking about a diverse range of methods they use to simulate their phenomena of interest. Presentations will be hosted on our
YouTube channel, at the discretion of the presenters.
Ann-Kristin Koehler is a Research Fellow in the Leeds School of Earth and Environment who has worked on crop modelling.
The General Large Area Model for annual crops (GLAM)
Process-based crop models are used to simulate the impact of weather and management on crop growth and yield. Traditionally designed for the field scale, the General Large Area Model for annual crops (GLAM) has been designed to operate on the spatial scales of global and regional climate models, i.e. much larger than field scale. Its complexity is low compared to field scale crop models, avoiding the need for large amounts of location-specific input data or calibration and is less computationally intensive allowing for large number of simulations aiding the quantification of uncertainties. It aims to simulate the impact of current sub-seasonal and inter-annual climate variability on observed yields.
Next to the impact of weather on yields, two other factors are simulated explicitly: planting date and soil type. Other factors impacting yields like pests, diseases and management factors are not modelled explicitly but are modelled using a single yield-gap parameter. Understanding the impacts of current climate variability on crop yields one can study the impacts of climate change on yields. If current trends of population growth and food consumption continue, global food production has to increase substantially.
Crop models like GLAM can contribute estimates and uncertainties of future global food production to inform adaptation strategies.
Leeds-Bradford Annual General Meeting
We will report on the year's activities and elect members of the committee for the coming year.
Verity Tether is a Doctoral researcher in the Leeds School of Geography and has used agent-based modelling to investigate crime patterns.
Using agent-based models to investigate patterns around crime clusters
It is a fact that crime occurs neither randomly nor uniformly in geographic space. Past research has identified two different types of crime clusters, crime generators and crime attractors, which emerge from differing causal mechanisms. At crime generators the majority of crimes are opportunistic; potential offenders encounter, and take advantage of, increased criminal opportunity due to the presence of large numbers of people. In the case of crime attractors, on the other hand, motivated offenders travel to these areas specifically to commit crimes, attracted by their reputation for criminal potential. In this research, agent-based models for these two types of spaces are created, to explore the crime patterns which emerge from each. This presentation will give an introduction to agent-based modelling, as well as explain how models were created to represent these spaces, before discussing their results and implications.
Tom Lawton is a Consultant in Critical Care & Anaesthesia and Head of Clinical Artificial Intelligence at Bradford Royal Infirmary who has modelled patient flows using discrete event simulation.
A whole-hospital model to investigate responses to Winter Pressure, in R
Making any change in an organisation as large and complex as the NHS is always difficult and potentially risky. Computerised modelling techniques offer the opportunity to obtain robust estimates of how effective any change might be. However, modelling is an under-used tool operationally in the NHS due to difficulties with using traditional techniques and a lack of data in the depth required to drive these.
Our work started by using a novel technique to model a single ward (the intensive care unit). We will show how we have expanded this to model an entire hospital, still using routinely collected data. Further we will demonstrate the use of this technique, combined with the Monte-Carlo method, to investigate patient flow around the difficult topic of responses to winter pressures.
We will show how we have used R to build this model, leveraging its power in data analytics to process data for model building, and its flexibility and expandability with the use of r-simmer and Rcpp to provide high performance for computationally-intensive and complex modelling.
Ann-Kristin Koehler is a Research Fellow in the Leeds School of Earth and Environment who has worked on crop modelling.
Verity Tether is a Doctoral researcher in the Leeds School of Geography and has used agent-based modelling to investigate crime patterns.
Tom Lawton is a Consultant in Critical Care & Anaesthesia and Head of Clinical Artificial Intelligence at Bradford Royal Infirmary who has modelled patient flows using discrete event simulation.
Organised by RSS Leeds-Bradford local group.