本文利用一維物理耦合生地化模式(王，2007)加上碳循環副程式(Ocean Carbon-Cycle Model Intercomparison Project; OCMIP)來模擬並探討南海SEATS(South-East Asian Time-series Study)測站表水二氧化碳分壓的時序變化。 透過兩種不同季節之初始場並藉著數位模擬結果：利用3月與9月間船測資料之平均做為初始場，來進行生地化與碳循環模式的模擬。在模式中，鹼度與溶解性無機碳的觀測值受物理及生地化作用驅動後，再透過OCMIP所提供的副程式，推估南海SEATS測站表水的CO2分壓，以決定是二氧化碳的源(source)或匯(sink)，並討論冬夏兩季海氣二氧化碳通量的大小與生物作用對南海SEATS測站的重要性。 得到模式模擬結果，並用觀測值修正後，海氣二氧化碳通量在冬季時為 -1.30 molC/m2/yr為二氧化碳的匯(sink)，在夏季海氣二氧化碳通量為1.21 molC/m2/yr ，為二氧化碳的源(source)；將全年度平均起來為0.07 molC/m2/yr，與前人估算資料符合(Chou, 2005)。 若無生物作用，則南海SEATS測站由海至氣二氧化碳通量在冬天將會增加0.88 molC/m2/yr，夏天則增加0.18 molC/m2/yr。 In this study, a one-dimensional model based on the Mellor and Yamada level 2.5 turbulence closure model was coupled to a biogeochemical model with carbon cycling to investigate the variation of pCO2 at the South-East Asian Time-series Study (SEATS) Site in the northern South China Sea (SCS) from 1997 to 2003. This study was focused on the biological effect on pCO2 and air-sea exchange of CO2. The model was initiated by two different initial conditions, one for March and the other for September. Both conditions are taken from averaged ship board measurements of the hydrographic and biogeochemical data, including alkalinity and dissolved inorganic carbon (DIC) data. The alkalinity and DIC concentrations, which are controlled by physical as well as biogeochemical processes, are used to calculate pCO2 values by the subroutine provided by OCMIP. The pCO2 data are used to determine whether the SEATS station is a carbon dioxide source or sink in different seasons.. The modeled sea to air fluxes of CO2 show a seasonal pattern similar to observations but differ somewhat in magnitudes. After calibrating the modeled fluxes with the observed values, we obtain the sea-to-air carbon dioxide flux to be -1.30 molC/m2/yr for winter and 1.21 molC/m2/yr for summer. The entire year average is 0.07 molC/m2/yr, close to Chou’s estimate (Chou, 2005). Without biology, the sea-to-air carbon dioxide flux will be higher by 0.88 molC/m2/yr in winter and 0.18 molC/m2/yr in summer.
