Analysis of Atmospheric Potential Oxygen (APO) in the Pacific region to evaluate the seasonal air-sea gas exchange

Yasunori Tohjima, NIES National Institute of Environmental Studies, Ibaraki, Japan, Hitoshi Mukai, National Institute of Environmental Studies, Tsukuba, Ibaraki, Japan, Toshinobu Machida, Natl Inst Environal Studies, Tsukuba, Japan, Shin-ichiro Nakaoka, CGER/NIES, Tsukuba, Japan, Ishizawa Misa, Misa.Ishizawa@Canada.Ca, QC, Canada and Tomoko Shirai, NIES, Tsukuba, Japan
Abstract:
Atmospheric Potential Oxygen (APO), which is defined by an equation: APO=1.1×CO2 + O2, is a unique tracer to study air-sea gas exchanges. This is because the air-sea exchanges of O2 and CO2 are main drivers of APO variation while APO is invariable with respective to land biotic processes. In general, APO increases during spring and summer and decreases during fall and winter. Such seasonal variations mainly reflect O2 outgassing associated with the ocean primary production and O2 ingassing associated with the ocean ventilation. To investigate the spatio-temporal variations in APO, we have been collecting air samples from several ground sites and commercial cargo ships sailing in the western Pacific and the northern North Pacific regions. The flask samples were sent back to our laboratory to measure the CO2 and O2 concentrations, then the APO is calculated. The APO data are binned into 10°×10° cells and the average seasonal cycles of the binned data are determined. Then the observed seasonal cycles are compared with simulated ones based on climatological ocean O2 fluxes prepared by Garcia and Keeling (2001) and an atmospheric transport model (NIES-TM 08i). The model simulation well reconstructs the observed APO seasonal cycles. However, there are consistent discrepancies in the seasonal variations in fall or winter: the observed decreasing rates during fall and winter are slower than the model simulations. The differences in the APO seasonal cycle (observation − simulation) show significant peaks in fall or winter. The amplitude and timing of the APO peaks increase and advance with latitude, respectively, in both hemispheres within the range of observations (40°S~50°N). The discrepancy between observed and simulated APO in fall may be attributed to the additional oceanic O2 emissions induced by the disappearance of the subsurface shallow oxygen maximum (SOM) and/or the fall blooming. The APO data could be used to improve the ocean O2 emissions in fall and winter.