Geospatial Modeler, Oak Ridge National Lab
Morgan is a Spring 2023 graduate who earned her Master of Science at the University of Tennessee Geography Department. She moved to Tennessee in 2017 to attend UTK for her Bachelor of Arts in Geography. Her graduate thesis work was inspired by her research position with Dr. Kelsey Ellis in undergrad. She helped classify storms by looking at radar images and identifying them based on their physical properties. She found the process subjective and time consuming, and she wanted to know more about automated approaches to storm classification, but she didn’t feel she had the skills to do so.
Between undergrad and grad in 2020, she interned at Oak Ridge National Laboratory (ORNL) within the Terrestrial Systems Modeling Group for nearly a year. Her group used a machine learning model to categorize tropical vegetation phenology using high resolution Sentinel-2 satellite data. She used Google Earth Engine to preprocess imagery on the cloud, and she explored open-source methods for analyzing and visualizing large datasets. She became familiar with Python and Bash, and it’s here that she discovered the great utility of unsupervised machine learning methods and cloud computing.
With newfound confidence, she returned to UTK to understand how storms are classified using machine learning algorithms. Specifically, she wanted to explore storm classification using unsupervised machine learning techniques, which are methods that don’t assume known categories but instead rely on a deep familiarity with the underlying data. To improve her data analysis and coding abilities, she interned as a data scientist at Maxar Technologies in 2022, where she worked on projects with an emphasis on big data synthesis, cloud computing, and geovisualization.
This culminated in her thesis work titled Unsupervised Machine Learning of Tornado-Producing Storms in the Southeast, which is currently being adapted for publication. Presently, she is looking to return to ORNL by the end of Summer 2023. For more information about Morgan and her past projects, see the links below.
Morgan’s Google Scholar