Premium
Inferring Air‐Sea Carbon Dioxide Transfer Velocities From Sea Surface Scatterometer Measurements
Author(s) -
Ghobadian M.,
Stammer D.
Publication year - 2019
Publication title -
journal of geophysical research: oceans
Language(s) - English
Resource type - Journals
eISSN - 2169-9291
pISSN - 2169-9275
DOI - 10.1029/2019jc014982
Subject(s) - scatterometer , middle latitudes , backscatter (email) , flux (metallurgy) , environmental science , wind speed , latitude , atmospheric sciences , buoy , radar , remote sensing , geology , meteorology , geodesy , geography , materials science , oceanography , telecommunications , computer science , metallurgy , wireless
C‐band microwave sea surface radar backscatter observations from the FINO‐2 tower in the western Baltic are analyzed with respect to their relevance for air‐sea CO 2 transfer velocity parameterizations. The scatterometer measurements observed from a height of 25 m above the sea surface using a multifrequency scatterometer instrument of the University of Hamburg were obtained quasi‐simultaneously with eddy covariance CO 2 flux measurements. Both data sets are merged here to derive a gas transfer velocity parameterization based on radar backscatter measurements. At the location of the FINO‐2 tower, the resulting time‐averaged gas transfer velocity amounts to 26.95 cm/hr. In combination with Δ PCO 2measurements available from the vicinity of the FINO‐2 platform, a time‐mean CO 2 flux of 0.23 μmol·m −2 ·s −1 into the Baltic was estimated. Applied to monthly mean satellite‐based C‐band ASCAT scatterometer data, the newly derived gas transfer velocity parameterization provides estimates of seasonal and annual mean global maps of air‐sea transfer velocities. The new results agree in their general pattern with previous estimates using wind speed parameterization. However, the backscatter‐based transfer velocities appear smaller at higher latitudes. Globally averaged air‐sea CO 2 fluxes would thereby be reduced by 20%. To what extent this is a robust result, or if it depends on the fact that the training data set did not represent conditions, has to be investigated in the future.