中文摘要：

文章综合分析了南海西部2012年秋季海表二氧化碳分压（p（CO2））分布及控制因素,估算并讨论了海-气CO2通量及其区域性差异。结果表明,秋季南海西部海表p（CO2）变化范围为37.8~57.1Pa,均高于大气p（CO2）,表现为大气CO2的源。研究海域海表p（CO2）分布主要受海水碳酸盐热力学的影响,海表温度（SST）是影响海表p（CO2）分布的主要因素。但在局部海域,海表p（CO2）分布还受到其他因素的控制。其中,越南东部局部海域表层海水受湄公河冲淡水的影响呈低盐特征,高p（CO2）的冲淡水团和离岸水团的混合作用直接影响该海域海表p（CO2）分布。此外,珊瑚礁的钙化活动造成南沙群岛弹丸礁西部海域出现表层p（CO2）极高值。秋季南海西部表层海水CO2释放速率的分布格局主要受风速的控制。北部海域海-气CO2分压差（Δp（CO2））低于中南部海盆,但其海表风速远高于中南部海盆,造成北部海域表层海水CO2释放速率（2.9mmol C·m^-2·d^-1）明显高于中南部海盆（0.9mmol C·m^-2·d^-1）。海-气CO2通量估算中不能忽视局部海域珊瑚礁代谢作用的影响。

英文摘要：

Distribution of p（CO2） in the surface water of the western South China Sea（SCS） was examined using data for autumn 2012.In this study,the control factors of surface p（CO2） were investigated and sea-air CO2 flux in the study area was estimated.The results showed that sea surface p（CO2） values of 37.8 to 57.1 Pa were observed in the western SCS.The sea surface p（CO2） was always higher than that in the atmosphere,which indicated the western SCS acted as a source of the atmosphere CO2 in autumn.In the study area,sea surface p（CO2） was mainly controlled by carbonate thermodynamics,and sea surface temperature（SST） was the predominant factor that controlled p（CO2） distribution.However,sea surface p（CO2） was also influenced by other factors in some areas.The transport of Mekong River diluted water led to low surface salinity water located in the Vietnamese eastern offshore region.Water mixing between Mekong River diluted water with high p（CO2） and offshore water affected the regional surface p（CO2） distribution directly.Furthermore,calcification process of coral reef organisms caused extremely high sea surface p（CO2） to exist in the western Swallow Reef of the Nansha Islands.The CO2 outgassing rate of the western SCS in autumn was mainly influenced by sea surface wind speed,with average CO2 outgassing rate of 2.9 mmol C·m^-2·d^-1 in the northern continental shelf and northern basin and of 0.9 mmol C·m^-2·d^-1 in the central and southern basins.The effect of metabolism of coral reefs in regional areas should not be neglected in CO2 flux estimation.