Publication Date:
2021-04-22
Description:
The seasonal cycle represents one of the largest signals of dissolved inorganic carbon (DIC) in the ocean, yet these seasonal variations are not well established at a global scale. Here, we present the Mapped Observation‐Based Oceanic DIC (MOBO‐DIC) product, a monthly DIC climatology developed based on the DIC measurements from GLODAPv2.2019 and a two‐step neural network method to interpolate and map the measurements. MOBO‐DIC extends from the surface down to 2,000 m and from 65°N to 65°S. We find the largest seasonal amplitudes of surface DIC in the northern high‐latitude Pacific (∼30 to 〉50 μmol kg−1). Surface DIC maxima occur in hemispheric spring and minima in fall, driven by the input of DIC into the upper ocean by mixing during winter, and net community production (NCP) driven drawdown of DIC over summer. The seasonal pattern seen at the surface extends to a nodal depth of 〈50 m in the tropics and several hundred meters in the subtropics. Below the nodal depth, the seasonal cycle of DIC has the opposite phase, primarily owing to the seasonal accumulation of DIC stemming from the remineralization of sinking organic matter. The well‐captured seasonal drawdown of DIC in the mid‐latitudes (23° to 65°) allows us to estimate the spring‐to‐fall NCP in this region. We find a spatially relatively uniform spring‐to‐fall NCP of 1.9 ± 1.3 mol C m−2 yr−1, which sums to 3.9 ± 2.7 Pg C yr−1 over this region. This corresponds to a global spring‐to‐fall NCP of 8.2 ± 5.6 Pg C yr−1.
Description:
Key Points
We present a near‐global monthly DIC climatology (MOBO‐DIC) based on ship observations and a two‐step neural network
Seasonal surface DIC amplitudes range from 0 to more than 50 μmol kg−1
MOBO‐DIC yields a spring‐to‐fall NCP in the euphotic zone of the mid‐latitudes of 3.9 ± 2.7 Pg C yr−1
Description:
European Community's Horizon 2020 Project
Description:
International Max Planck Research School on Earth System Modelling (IMPRS‐ESM)
Keywords:
551
;
DIC
;
seasonal variability
;
neural networks
;
SOM‐FFN
;
monthly climatology
;
NCP
Type:
article
Permalink
