Accurate estimations of ocean primary production by phytoplankton at large spatial and temporal scales are essential to determine variations in marine biogeochemical cycles and ecosystems. Consequently, satellite ocean color remote sensing has been used to estimate primary productivity on a global scale. This paper reviews the development of models for primary productivity retrieval from satellite data and summarizes the available literature on long-term variations in ocean color chlorophyll and primary production. Although chlorophyll-based models are often applied to estimate primary productivity, carbon- and light absorption-based models were developed recently to reduce errors due to satellite chlorophyll a data attributed to packaging effect and the influence of suspended and colored dissolved organic matter. Furthermore, correcting drift in ocean color data and gaps among ocean color satellite missions is necessary to achieve adequate evaluations of long-term variations in ocean color chlorophyll and primary production.

The subarctic North Pacific Ocean is a high-nutrient, low-chlorophyll area in which nutrients remain undepleted in surface waters throughout the year. In this area, phytoplankton growth is limited by the availability of Fe in seawater. To improve our understanding of primary production in the subarctic North Pacific Ocean, it is important to determine Fe distributions and supply processes for this area. However, Fe distribution and biogeochemical cycling have not been fully clarified, because Fe is a contamination-prone trace metal during onboard sampling and analyses in the laboratory. In this review, we compile information on Fe distribution and supply processes in the subarctic North Pacific Ocean from previous studies, with a focus on the atmospheric transportation of aerosols from continents and lateral transportation from coastal areas as Fe supply processes. Moreover, we discuss future studies that should be undertaken to address knowledge gaps in this research topic.

Iron is an important nutrient in marine biogeochemical cycles because it limits primary production in high-nutrient, low-chlorophyll areas and nitrogen fixation rates in nitrogen-limited areas. Although three-dimensional biogeochemical cycle models were developed in the 1990s to evaluate oceanic carbon dioxide uptake, it was not until 2000 that such models considered iron cycling. This was mainly because of limited field data on iron, arising from difficulties in measuring dissolved iron. Early iron cycle models were developed from knowledge based on limited field data, and such models assumed that global-scale dissolved iron distribution was determined simply by iron supply from aeolian deposition and removal by scavenging via complexation with organic ligands. Subsequent accumulation of field data has revealed that various external sources, cycling of organic ligands, and colloidal aggregation can affect global- and basin-scale distributions of dissolved iron. In this paper, we review the history of the development of marine biogeochemical cycle models with a special focus on iron, and describe their current status and issues that must be addressed for further model improvement. Recent activity regarding the Iron Model Intercomparison Project is also described.