DOI: 10.1029/2025gb008952 ISSN: 0886-6236

CbPM Estimates of Net Primary Production in the North Atlantic From Profiling Floats and Satellites Diverge Seasonally Due To Fluorescence and Vertical Extrapolation Effects

Nina Buzby, Andrea J. Fassbender, Alison R. Gray, Marin Cornec, Jacki Long, Ellen Park

Abstract

Marine net primary production (NPP), defined as the difference between gross production and phytoplankton respiration, is often estimated using algorithms applied to remote sensing data. While assumptions are needed to extend surface satellite observations through depth, some NPP algorithms, like the Carbon‐based Productivity Model (CbPM), have been adapted to vertically resolved data collected by autonomous profiling floats. Such applications eliminate the need for vertical extrapolation but introduce challenges related to float measurements of fluorescence rather than chlorophyll‐ a (Chl‐ a ; required CbPM input). This study analyzes over a decade of float observations from the North Atlantic to estimate NPP using CbPM and quantify its sensitivity to different input parameters: (a) fluorescence versus Chl‐ a , (b) vertically extrapolated versus depth‐resolved information, and (c) in situ versus remote observations of the first optical depth—the impacts of which vary seasonally and regionally. In higher latitude waters, converting float fluorescence to Chl‐ a using a novel correction based on satellite data produces significantly smaller NPP estimates at seasonal and annual timescales. In contrast, extrapolation and platform‐related differences largely compensate when integrated vertically and annually, such that cumulative annual depth‐integrated NPP (iNPP) estimates computed with fluorescence‐corrected float measurements are statistically indistinguishable from those extrapolated from satellite observations. These effects are reversed in the subtropics: discrepancies due to fluorescence compensate vertically and annually, whereas annual iNPP estimates from depth‐resolved float measurements significantly outweigh those of satellites. Seasonal changes to the sign, timing, and vertical structure of NPP discrepancies suggest persistent sub‐seasonal disagreement between platforms, highlighting knowledge gaps in understanding NPP.

More from our Archive