Remote sensing algorithm for sea surface CO<sub>2</sub> in the Baltic Sea

Abstract

Abstract. Studies of coastal seas in Europe have brought forth the high variability in the CO2 system. This high variability, generated by the complex mechanisms driving the CO2 fluxes makes their accurate estimation an arduous task. This is more pronounced in the Baltic Sea, where the mechanisms driving the fluxes have not been as highly detailed as in the open oceans. In adition, the joint availability of in-situ measurements of CO2 and of sea-surface satellite data is limited in the area. In this paper, a combination of two existing methods (Self-Organizing-Maps and Multiple Linear regression) is used to estimate ocean surface pCO2 in the Baltic Sea from remotely sensed surface temperature, chlorophyll, coloured dissolved organic matter, net primary production and mixed layer depth. The outputs of this research have an horizontal resolution of 4 km, and cover the period from 1998 to 2011. The reconstructed pCO2 values over the validation data set have a correlation of 0.93 with the in-situ measurements, and a root mean square error is of 38 μatm. The removal of any of the satellite parameters degraded this reconstruction of the CO2 flux, and we chose therefore to complete any missing data through statistical imputation. The CO2 maps produced by this method also provide a confidence level of the reconstruction at each grid point. The results obtained are encouraging given the sparsity of available data and we expect to be able to produce even more accurate reconstructions in the coming years, in view of the predicted acquisitions of new data.

Using satellite data to estimate partial pressure of CO2 in the Baltic Sea

“…The outputs of this research have a horizontal resolution of 4 km and cover the 1998–2011 period. We continue the work of Parard et al [] by estimating p CO 2 variability in time and space over the Baltic Sea.…”

“…To reconstruct the sea surface p CO 2 concentrations, we employed the SOMLO methodology [ Sasse et al , ] in a similar way to that applied by Parard et al []. The SOMLO methodology combines two statistical approaches: self‐organizing maps (SOMs) [ Kohonen , ] and linear regression .…”

“…Parard et al [] used the self‐organizing multiple linear output (SOMLO) [ Sasse et al , ] method to estimate the ocean surface p CO 2 in the Baltic Sea from the remotely sensed SST, Chl a , CDOM, net primary production (NPP), and mixed‐layer depth (MLD). The outputs of this research have a horizontal resolution of 4 km and cover the 1998–2011 period.…”

The potential of using remote sensing data to estimate air–sea CO<sub>2</sub> exchange in the Baltic Sea

“…1994 to 2008), the CB acts as a source of 1.64 mol m −2 yr −1 for the atmosphere. As we explain in Parard et al (2014), the pCO 2 data do not reproduce the spring-summer bloom in the eastern Gotland Sea described in Schneider et al (2015). The data used for the computation contain the voluntary observing ships (VOSs) ship line, but we calculated a monthly average.…”

Construction of Multi-Year Time-Series Profiles of Suspended Particulate Inorganic Matter Concentrations Using Machine Learning Approach

“…Potential benefits of these methods include filling data gaps caused by cloud cover, inferring vertical distributions of biogeochemical variables observed at sea surface, emulate models of the upper ocean dynamics at high spatial resolution and developing relationships between environmental variables [9][10][11][12]. Further applications can be found in carbon dynamics [13][14][15], coastal turbidity [16,17], phytoplankton blooms [18][19][20][21], harmful algal blooms [22,23], and primary production [24].…”