Joe Guinness, assistant professor, Department of Biological Statistics and Computational Biology
Academic focus: My research is focused on analyzing large spatial-temporal datasets in Earth and environmental sciences, including interpolation of data from weather stations, managing and making sense of large climate simulation datasets, and learning about our planet using data from satellites. This work involves theoretical understanding of new statistical models, and development of computationally efficient algorithms and software.
Previous Positions: assistant professor, North Carolina State University Statistics, 2014-18; visiting assistant professor, Cornell University, 2017-18; postdoctoral scholar, North Carolina State University, Statistics, 2012-14
Academic Background: B.A., mathematics and physics, Washington University in St. Louis, 2007; Ph.D. statistics, University of Chicago, 2012
Last Book Read: The Undoing Project by Michael Lewis
What you do when not working? I do a lot of running. You can find me in Stewart Park avoiding the hills.
What gets you out of bed in the morning? Working with students and collaborating with scientists. I also enjoy writing and optimizing code.
Current research projects? Inferring wind patterns from geostationary satellite data, predicting lightning flash rates from multispectral cloud images, estimating pollution exposure rates from air quality monitors, general computational methods for working with a statistical model called a Gaussian process.
Current outreach/extension projects? I am a faculty facilitator for the Math Alliance program, which provides guidance and resources for undergraduate students from backgrounds that are underrepresented in the mathematical sciences, as they apply for graduate programs.
What are three adjectives people might use to describe you? Statistically speaking, there’s got to be somebody out there who would describe me as handsome, charming, and smart. That person is probably my mother.
Course you’re most looking forward to teaching? Biometry 6020, a second statistics course for non-statistics graduate students. I taught this course last spring when I visited and had a blast. I consider teaching this class to be a very important and impactful part of my job. It’s an opportunity to equip emerging scientists with the knowledge to conduct high quality, rigorous statistical analyses.
If you had unlimited grant funding, what major problem in your field would you want to solve? If it were truly unlimited, I would do a massive experiment to see if building renewable energy infrastructure all over the world would reduce atmospheric carbon dioxide and curb global warming. If it were just a lot of money, I would study data compression and decompression for spatial-temporal datasets. It’s not the most glamorous problem, but the cost of storing satellite data and numerical model output comes up over and over again. In practice, data storage is a bottleneck for many research projects in the Earth and environmental sciences.
What most excites you about Cornell CALS? Definitely the opportunity to contribute to the digital agriculture initiative. Cornell CALS will be a leader in digital agriculture, so it is exciting to be a part of it.