Please join us next Friday, March 7, at 12 pm for a Geochemistry Seminar by LDEO alum Dr. Rik Wanninkhof (Ph.D. ’86), a senior scientist at NOAA and Atlantic Oceanographic and Meteorological Laboratory. This is a special Lamont75 Seminar, where eminent alums and former staff showcase their research and reflect on what makes LDEO special.

Estimates of global sea-air CO2 fluxes: thoughts on observational versus observation-based products by Rik Wanninkhof with contributions from Joaquin Triñanes, Peter Landschutzer, Amanda Fay, David Munrol, Colm Sweeney, and Annika Jersild.

Rik will describe progress in quantifying global sea-air CO2 fluxes based on observations. The presentation will provide a historical perspective on the overall construct of the estimates that is based on seminal contributions by the late Wally Broecker and Taro Takahashi of LDEO. The advances from monthly resolved climatologies of surface water pCO2 levels and fluxes to our current (sub)monthly estimates on a (sub)1-degree grid have been accomplished through incorporating a wealth of global ocean data heavily relying on remote-sensed products utilizing machine learning (ML) approaches. Different observation-based estimates yield similar results, increasing confidence in these approaches. However, part of the correspondence appears to be related to using similar input parameters and an over-constrained system with correlated predictor variables. The effect of changing predictor and target variables, along with a discussion of other factors other than pCO2, will be illustrated following the recent paper by Wanninkhof et al. (2025).

Wanninkhof, R., Triñanes, J., Pierrot, D., Munro, D. R., Sweeney, C., & Fay, A. R. (2025). Trends in sea-air CO2 fluxes and sensitivities to atmospheric forcing using an extremely randomized trees machine learning approach. Global Biogeochemical Cycles, 39(e2024GB008315). doi:https://doi.org/10.1029/2024GB008315