LG.02 - School of Mathematics, Fry Building, Woodland Road 

This event aims to highlight the real-world applications of mathematics to sustainability related problems. Led by PhD students, a series of talks will display ongoing research already taking place within our department. Following each presentation, attendees will have the opportunity to engage in a Q&A session, and find out more about their research.

Speakers:

Cecina Babich Morrow - Bristol

Title: Sensitivity of Bayesian Decision Analysis: a tool for robust climate adaptation decision making

Abstract: Bayesian Decision Analysis (BDA) is a framework for decision-making given an uncertain state of nature, making it a powerful tool for climate change adaptation decision-making. These decisions rely on the estimation of risk, which is modeled using information about hazard (the source of potential damage), exposure (the amount of damage experienced), and vulnerability (the level of susceptibility to damage), each of which has a high degree of uncertainty. We apply BDA to an idealised example application of adaptation options to combat the effects of heat-stress. Previous work has investigated the sensitivity of BDA to variations in hazard, exposure, and vulnerability. We build on these analyses to investigate how variations in the attributes of individual adaptation options, such as costs, affect the decision outcome, identifying the most influential inputs and investigating how these vary spatially. Understanding which factors have the greatest influence on the optimal decision is crucial for transparent and robust climate adaptation decision-making.

Dan Milner - Bristol

Title: What Builds Climate Resilience Among Smallholder Farmers in Kenya and Ethiopia?

Abstract: Climate change poses a significant threat to the livelihoods of smallholder farmers in East Africa, exacerbating food insecurity and disproportionately affecting vulnerable communities. To address this pressing issue, enhancing farmers' resilience to future climate impacts emerges as a crucial solution. This presentation explores the potential of leveraging multiple sources of data to deepen our understanding of farming practices, the challenges posed by climate change, and smallholder farmers' responses within two distinct study sites. By synthesizing data from various sources, including open source data, satellite imagery, and socioeconomic surveys, we aim to gain comprehensive insights into the dynamics of resilience in the face of climate variability.

One key aspect highlighted is the pivotal role of mathematics in navigating the complexity of the available data, particularly in environments characterised by high levels of data noise. Mathematical modelling and analysis techniques provide indispensable tools for extracting meaningful patterns, identifying trends, and elucidating the relationships between climatic variables, agricultural practices, and socio-economic factors.

Sam Stockman - Brisol

Title: Modelling earthquakes and their aftershocks with point processes 

Abstract: Earthquakes are the natural hazard that cause the most fatalities globally, yet decades of study into the physics of plate tectonics has not resulted in a great deal of earthquake predictability. Statistical models that consider earthquakes as points in time and space have proven to be a more useful tool for forecasting, including being able to better express our uncertainty about the earthquakes yet to happen. I will outline how maths and statistics are being used by organisations around the world in earthquake forecasting models, as well as what research is being conducted to improve these models.

Shannon Williams - Bristol

Title: Quantifying and incorporating uncertainty in probabilistic volcanic ash hazard forecasting

Abstract: In the event of an explosive volcanic eruption, volcanic ash is dispersed through the atmosphere by meteorological processes and eventually deposited on the ground, possibly hundreds of kilometres away from the eruption. Airborne ash can impact aviation, causing aircraft to be grounded or re-routed to prevent damage to their engines. Downwind deposition can lead to a range of further impacts, including to infrastructure, agriculture and populations. These processes are dynamic and highly uncertain, and are modelled by atmospheric dispersion simulators which are highly dependent on their inputs (meteorological data and eruption source parameters). Commonly an ensemble of simulator outputs is constructed and aggregated to capture the uncertainty, however there is little consensus in the volcanological community on how to construct and interpret these ensembles. I will briefly present two pieces of work designed to address the problem of incorporating and quantifying uncertainty in probabilistic volcanic ash hazard forecasting in both the long and short term.

Register here

Light refreshments will be served in the Fry Atrium after the event, this social gathering will incorporate the opportunity to ask our speakers any questions you may have.