In my recent work, I have used neural networks to improve a vertical mixing scheme in the ocean component of a climate model and it has led to a better representation of physics along with a reduction of biases. My article shows how data-driven approaches can replace missing physics in Earth System Modeling under a pardigm known as ‘hybrid modeling’. The article can be found here.

My doctoral studies focused on a regional model of an estuary: the Narragansett Bay. I showed that a model of the Narragansett Bay and surrounding regions can be used for forecasting future conditions like temperature, currents, etc. Check out my paper here. This work forms a basis for establishing an oceanic forecasting system for the Bay region of Rhode Island.

Prior to being an oceanographer, I studied interfacial fluid dynamics. I used theory and experiments to measure the surface tension of flowing soap films. We found a previously unknown dependence of the soap film’s terminal velocity on the film thickness and its flow rate. You can find my article here and a commentary paper on our work can be found here.