R code for global reconstruction and uncertainty analysis of phytoplankton dissolved primary production

Le résumé fourni par la source

This repository contains the R code used for the analyses presented in Huang et al. (submitted). The scripts mianly include development, comparison, and validation of Random Forest models for the percentange of phytoplankton extracelluar release (PER), including tests of spatial autocorrelation, alternative phytoplankton community predictors, and Monte Carlo uncertainty analyses. The trained models are used to reconstruct global and seasonal distributions of PER and to estimate global dissolved primary production (DPP) and associated carbon fluxes. The code further implements first-order scaling of DPP-derived refractory dissolved organic carbon (RDOC) formation, parameter sensitivity analyses, and evaluation of its potential contribution to long-term carbon sequestration relative to the particulate pathway.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.