We are delighted to welcome Dr Teresa A’mar to Dragonfly. She has an impressive background in fisheries science, software development and statistical modelling.
Teresa says she first encountered Dragonfly while working as a fisheries modeller at NIWA. She was involved in a simulation modelling project with Phil Neubauer and NIWA principal fisheries scientist Matt Dunn.
“It was an enjoyable collaboration. I was impressed by how sharp and collegial Phil was. Phil, Matt, and I were able to communicate well, explore a range of ideas and implementations, and get a lot of things done.”
Teresa also worked with Kath Large at NIWA, mainly on the first release of the Casal2 population modelling software package.
When Kath took up a senior fisheries scientist role with Dragonfly in 2021, Teresa watched her progress with interest. “I was curious about her move, and it was clear that Kath was thriving at Dragonfly.”
Teresa was also a co-supervisor for Kyuhan Kim’s PhD thesis, along with academic staff from Victoria University of Wellington.
“I was struck by how focused and self-motivated he was. I was glad to recommend him to Dragonfly for his current role, and I’m pleased to see him developing into a well-rounded scientist.”
The expertise sharing went both ways. Teresa has benefitted from Kyuhan’s tutoring of the Korean language through the informal conversation sessions he leads.
Teresa moved to New Zealand in 2019 to take up the NIWA role. Her PhD in quantitative ecology and resource management is from the University of Washington in Seattle, supervised by Professor André Punt.
“André is renowned in fisheries science and has a diverse range of research interests and collaborators. He guided my research to develop a simulation framework to test the harvest strategy used for managing the walleye pollock stock in the Gulf of Alaska. A stock assessment and a harvest control rule are used to set the target catch level for the following year, and we projected the stock to see how the management strategy performed under a range of scenarios.”
“The projection process needed a lot of computing time. At one point I went out and bought a gaming computer in order to have dedicated computing resources. It took 42 sets of weeklong runs to get through all of the projections.”
After her studies, Teresa worked as an operations research analyst at NOAA Fisheries and as a statistical research analyst at the Fred Hutchinson Cancer Research Center in Seattle.
“The group I worked with at Fred Hutch was doing some really interesting projects. The one I worked on most recently was detecting if and when breast cancer had recurred. We used machine learning on patient diagnosis information, cancer registry data and insurance claims.”
“Unfortunately our dataset was too small to have robust results but we helped another research group with more data to apply the data processing and analytical methods we’d developed. I spent time working with them on exploring and preparing their data and transferring the machine learning methods and code – I am really curious to see how they got on.”
Teresa says working with data and models has taught her how to help people plan and structure their work environments, rather than having to argue with the components and interfaces. It’s something that’s also part of her role at Dragonfly.
“I want to make it easier for others by documenting our data and modelling processes and guidelines, and generalising existing frameworks. When things are set up well, it’s more efficient because you don’t have to start every project from scratch.”