Surface Ocean PCO2 from the Surface Carbon State Estimate (SCSE)
D. Munro1,2, A.R. Jacobson1,2, A. Fay3, P. Landschützer4, C. Rödenbeck5 and C. Sweeney2
1Cooperative Institute for Research in Environmental Sciences (CIRES), University of Colorado, Boulder, CO 80309; 303-497-5765, E-mail: firstname.lastname@example.org
2NOAA Global Monitoring Laboratory (GML), Boulder, CO 80305
3Department of Earth and Environmental Sciences, Columbia University and Lamont Doherty Earth Observatory
4The Ocean in the Earth System, Max Planck Institute for Meteorology
5Biogeochemical Signals, Max Planck Institute for Biogeochemistry
We present results from the Surface Carbon State Estimate (SCSE), a Kalman filter interpolation scheme for mapping surface ocean pCO2 over the global oceans. This method was designed to provide a statistically well-characterized prior estimate for air-sea CO2 fluxes within CarbonTracker but also fills an important need emphasized within the surface ocean pCO2 community, namely, the need to develop pCO2 mapping methods with realistic time-varying estimates of uncertainty. SCSE tracks the magnitude of a set of basis functions over time; in this analysis, we focus on empirical orthogonal functions (EOFs) of surface ocean pCO2 output from a set of CMIP5 ocean biogeochemical simulations. Unlike other pCO2 mapping approaches, uncertainties are explicitly characterized by a full-rank posterior covariance matrix and respond at each time step to the spatiotemporal coverage of observations. We tested SCSE in an OSSE framework where we use the MPI-SOMFFN pCO2 product [Landschützer et al. 2014] as a truth condition and show that the basis functions derived from CMIP5 model output are adequate to reconstruct surface ocean pCO2. Additionally, we show that SCSE's uncertainty estimates are accurate. We use SCSE and the SOCAT database to investigate long-term trends in global and regional air-sea CO2 fluxes from the 1980’s to the present. We compare results from SCSE with other widely-used time-varying surface ocean pCO2 products [Fay et al. 2021] and investigate how seasonal biases in existing sampling networks influence understanding of long-term change in the seasonal cyle of surface ocean pCO2 and air-sea CO2 fluxes in several important regions including the subtropical North Pacific and the Southern Ocean.
Figure 1. Comparison of surface ocean ΔpCO2 (surface ocean pCO2 minus atmospheric pCO2) estimates for a biome in the subtropical North Pacific [Fay and McKinley 2014] based on surface ocean pCO2 results from the Surface Carbon State Estimate (SCSE) (black) with time-varying uncertainty (gray shading), a pCO2 reference based on the climatology of Takahashi et al.  and application of a constant trend (red), the Jena-MLS [Rödenbeck et al. 2013] (cyan) and MPI-SOMFFN [Landschützer et al. 2014] (blue).