School of Mathematics

Ken Newman

Lochlainn Partis-Marsh has written the following article as part of our series of Academic Interviews; featuring Ken Newman!

Dr. Newman started out his career in ecology, specifically completing a bachelor’s degree in natural resources; he would then go on to achieve a masters in forest management. Over the course of his undergraduate and postgraduate studies however, he became more and more interested in statistics; partly due to its generality. From there he decided to fully embrace statistics, studying it for a masters at Oregon State University. He returned to his ecology roots for several years after this; working for the Northwest Indian Fisheries Commission in Washington State, before completing a PhD in statistics in 1993 from the University of Washington.

Dr. Newman would join the US Fisheries and Wildlife service as a mathematical statistician in 2006. It was here that he would use statistics and statistical models to affect decisions which would make real impacts on the world. Dr. Newman’s work constructing statistical models would help make, and justify, management decisions which would impact the environment. One particular example he mentioned was his work in a court case involving a small fish called the Delta Smelt. The issues revolved around the use of river water for irrigation purposes, the Judge in the case requested that a model be made to show and explain what impact the potential draining, or use, of water in California rivers would have on the species. In particular a life cycle model was constructed; an example of a state space model. This presented a microcosm of his typical work. He first had to consider and assess the various sampling methods and experiment designs to ensure his model was robust and rigorous. He also consulted and had discussions with biologists; considering how best to analyse and use the biological data that had been collected. The final phase of the study was the constructing and fitting of models to the data he had, and then analysing them. These models were then used to change, justify or defend decisions made with respect to the environment. In the case of the Delta Smelt the life cycle model was used to assess the impact of the use of river water for irrigation of farms. I myself was interested in his work here, as it presented an insight into how statistics can be used to help justify and influence decisions affecting the environment. At present, it is more important than ever to know how decisions we make could affect the environment, and assessing the best course of action to take to protect it.

Dr. Newman currently works as a principal researcher in statistical methodology within Biomathematics and Statistics Scotland (BIOSS), alongside his teaching role within the School of Mathematics. He praised both the School and BIOSS for being a welcoming and collaborative place to work. Deeply enjoying his current research, if also being slightly challenged by it.

A piece of Dr. Newman’s work I found particularly interesting, was his work with statistical emulators on crop growth and modelling phosphorous run-off from farms. Modelling crop growth and phosphorous run-off into streams and the local environment are both important things to know about, but also very complex things to understand. Dr. Newman said there are around 600 inputs and 600 outputs for the crop growth model, for example. This is where statistical emulators come in, they are essentially attempts to create a model that is far simpler to use and understand than other models, such as the one above. The statistical emulators Dr. Newman uses are a variation of multiple linear regression. This is a complicated way of saying you have a linear function which has numerous inputs, or “predictors”, which are the factors affecting how, for example, your oat crop grows, to then produce predictions; you take a “weighted” sum of these predictors to produce a prediction. We weight or adjust the predictors to ensure that the level of impact each factor has on the outcome is taken into account. Dr. Newman also described his work predicting the number of aphids in a given year. Using data that has been collected in the Edinburgh area for many decades, and combining that with weather data, allows you to predict the number of aphids we may get.

Interviewing Dr. Newman was a deeply interesting experience, especially hearing how different statistical methods are used within the context of the environment. This lets us fully understand and assess the impact of our decisions, and decide how to make better ones that protect the environment and benefit us all.