Using machine Learning to Improve Simulation of Phytoplankton Size Classes in Earth System Models

Wednesday, November 04, 2026 at 12:30pm to 1:30pm

SMAST East 101-103 : 836 S. Rodney French Boulevard, New Bedford MA 02744
Callie Rumbut
c.rumbut@umassd.edu
https://umassd.zoom.us/j/97440069270

Seminar Announcement

Department of Estuarine and Ocean Sciences

"Using machine Learning to Improve Simulation of Phytoplankton Size Classes in Earth System Models"

Sandupal Dutta

PhD Candidate, Dept. Earth and Planetary Sciences, Johns Hopkins University

Wednesday, November 4, 2026

12:30 - 1:30 pm

SMAST E 101-103 and via Zoom

Abstract:

Marine phytoplankton is responsible for fixing approximately 50 Gt C/yr and powers the biological pump, which is an important part of the carbon cycle. Phytoplankton are often classified according to their sizes. Earth System models (ESMs) such as the E3SM, NCAR CESM2, NASA GISS has significant biases in predictions of plankton size classes. We use satellite ocean color data from Copernicus and Scientific Machine Learning to alter the governing equation of plankton used by MARBL (Marine Biogeochemistry Library), biogeochemistry module of the ESMs, to improve the prediction of plankton size classes by ESMs.

Join Meeting

https://umassd.zoom.us/j/97440069270

Note: Meeting ID and passcode required. Email contact to obtain.

For additional information, please contact Callie Rumbut at c.rumbut@umassd.edu