Satellite-Based Assessment and Model Intercomparison of Ocean Net Primary Productivity (NPP) in the Northern Indian Ocean
摘要
Consistent data on net primary production (NPP) are essential for understanding of ecosystem responses to climate change and constraining carbon flux estimates in Earth system models. Although coordinated NPP observations have been conducted for more than three decades, comprehensive global syntheses remain limited. This study presents a global NPP database and a comparative evaluation of four satellite-based NPP models: the Carbon, Absorption, and Fluorescence Euphotic (CAFÉ) model; the Carbon-Based Production Model (CBPM); the chlorophyll-linked Eppley model; and the Vertically Generalized Production Model (VGPM). These models were assessed using MODIS-Aqua satellite inputs to characterize spatial and temporal patterns of ocean productivity at both global and regional scales, with particular focus on the Northern Indian Ocean, Arabian Sea, and Bay of Bengal. Model outputs for 2019 reveal monthly NPP values ranging from 380 to 1200 mg C m⁻² d⁻¹, with the Arabian Sea exhibiting the highest productivity. Among the models, VGPM demonstrated the strongest performance when validated against MODIS-Aqua-derived chlorophyll a, sea surface temperature (SST), and particulate organic carbon (POC) datasets for 2019–2021, while CAFÉ uniquely predicted elevated NPP across the Northern Indian Ocean. Validation with in situ carbon-13 uptake measurements confirmed the superior accuracy of VGPM relative to the other models. Long-term analyses from 2003 to 2021 indicate pronounced interannual variability, with the Arabian Sea showing peak NPP values of 1800–2200 mg C m⁻² d⁻¹, significantly higher than those of the Bay of Bengal, the broader Northern Indian Ocean, and the global ocean. Localized extreme NPP hotspots (23000–35000 mg C m− 2 d− 1) were also identified at the global scale during May-June, corresponding to phytoplankton bloom periods in temperate regions. Overall, these results demonstrate the value of satellite-based modeling, particularly VGPM, for capturing regional productivity dynamics and advancing global assessments of oceanic carbon cycling.
Graphical Abstract