Nutrients are typically the most important determinant of the productivity of marine ecosystems. Hence, nutrients have been an essential variable of ocean observations in the modern oceanography era. Understanding of marine ecosystems and biogeochemistry, however, is largely limited by the spatiotemporal coverage of nutrient data. Herein, we developed a novel algorithm based on a large observational dataset of nutrients and their relationship with water masses indexed by temperature and salinity in the South China Sea (SCS). The algorithm yielded errors of ≤ 1.3, ≤ 0.10 and ≤ 3.5 μmol L⁻¹ for NO₃^– + NO₂^– (N + N), phosphate and silicate, respectively. It is then applied to reconstruct nutrient concentrations primarily using temperature and salinity data archived in the World Ocean Database and Argo database during 1940–2018. It increases the nutrient data to ~ 5 million, by ca. three orders of magnitude compared to direct measurements in the SCS. This allows for a full examination of the seasonal climatology of nutrients in the SCS. In summer, in the upper 200 m, nutrient concentrations in the northwest and southernmost SCS are higher than in the rest of the SCS. An overall reversed pattern is revealed in winter, when higher nutrient concentrations are found in the central basin rather than basin margins. The Kuroshio intrusion and the vertical displacement of the nutricline, driven by upwelling/downwelling induced by horizontal convergences/divergences at both meso- and basin scales, determine the spatial and seasonal variation of nutrients. Seasonally, the Kuroshio intrusion and vertical displacements of the nutricline tend to offset the spatial variation of nutrient concentrations in spring, while they show an additive effect in the fall.

营养通常是海洋生态系统生产力的最重要决定因素。因此，营养物已成为现代海洋学时代海洋观测的重要变量。然而，对海洋生态系统和生物地球化学的了解在很大程度上受到营养数据时空覆盖的限制。本文中，我们基于大量的营养素观测数据集及其与南海（SCS）的温度和盐度指数所指示的水质之间的关系开发了一种新颖的算法。对于NO ₃ ^– + NO ₂ ^–，该算法产生的误差分别为≤1.3，≤0.10和≤3.5μmolL ^-1。（N + N），磷酸盐和硅酸盐。然后将其主要用于利用1940-2018年间在世界海洋数据库和Argo数据库中存储的温度和盐度数据重建营养物浓度。它将养分数据增加到约500万个。与SCS中的直接测量相比，三个数量级。这可以对SCS中养分的季节性气候进行全面检查。夏季，在200 m的上部，西北部和最南端的南半球的养分含量高于其余南半球。在冬季，当在中部盆地而不是盆地边缘发现较高的养分浓度时，揭示出总体上相反的模式。黑潮的入侵和营养素的垂直位移，在中尺度和盆地尺度上，由水平收敛/发散引起的上涌/下涌驱动，决定了养分的空间和季节变化。季节性上，黑潮的入侵和营养素的垂直位移往往会抵消春季营养物浓度的空间变化，而秋季则表现出累加效应。