The ocean annually absorbs about a quarter of all anthropogenic carbon dioxide (CO₂) emissions. Global estimates of air–sea CO₂ fluxes are typically based on bulk measurements of CO₂ in air and seawater and neglect the effects of vertical temperature gradients near the ocean surface. Theoretical and laboratory observations indicate that these gradients alter air–sea CO₂ fluxes, because the air–sea CO₂ concentration difference is highly temperature sensitive. However, in situ field evidence supporting their effect is so far lacking. Here we present independent direct air–sea CO₂ fluxes alongside indirect bulk fluxes collected along repeat transects in the Atlantic Ocean (50° N to 50° S) in 2018 and 2019. We find that accounting for vertical temperature gradients reduces the difference between direct and indirect fluxes from 0.19 mmol m⁻² d⁻¹ to 0.08 mmol m⁻² d⁻¹ (N = 148). This implies an increase in the Atlantic CO₂ sink of ~0.03 PgC yr⁻¹ (~7% of the Atlantic Ocean sink). These field results validate theoretical, modelling and observational-based efforts, all of which predicted that accounting for near-surface temperature gradients would increase estimates of global ocean CO₂ uptake. Accounting for this increased ocean uptake will probably require some revision to how global carbon budgets are quantified.

海洋每年吸收约四分之一的人为二氧化碳 （CO₂） 排放。空气-海洋 CO₂ 通量的全球估计通常基于空气和海水中 CO₂ 的大量测量，而忽略了海洋表面附近垂直温度梯度的影响。理论和实验室观测表明，这些梯度会改变空气-海洋_CO2 通量，因为空气-海洋_CO2 浓度差异对温度高度敏感。然而，到目前为止，缺乏支持其效果的现场证据。在这里，我们展示了 2018 年和 2019 年沿大西洋（北纬 50° 至 南纬 50°）沿重复样带收集的独立空气-海洋 CO₂ 通量以及间接体通量。我们发现，考虑垂直温度梯度可将直接和间接磁通量之间的差异从 0.19 mmol ^{m-2 d-1} 减少到 0.08 mmol ^{m-2 d-1} （N = 148）。这意味着大西洋 CO₂ 汇增加了 ~0.03 PgC ^yr-1（~7% 的大西洋汇）。这些实地结果验证了理论、建模和基于观测的努力，所有这些都预测考虑近地表温度梯度将增加对全球海洋 CO₂ 吸收的估计。要解释这种增加的海洋吸收量，可能需要对全球碳收支的量化方式进行一些修改。