2003年夏季東海表水二氧化碳之時空變化與控制機制探討

Abstract

本研究主要是探討2003年夏季東海表水二氧化碳分壓之時空分佈，控制機制與其海氣交換通量之變化。研究時間為2003年6月19日至26日與2003年8月13日至23日，於東海（East China Sea, ECS）進行現場偵測海水與大氣中之二氧化碳分壓（fCO2）。大氣fCO2於2003年6月與8月航次期間無明顯的變化，平均各自為365.9 ± 1.7與359.8 ± 3.6 μatm。而2003年6月觀測期間表水fCO2範圍介於 109.1 - 435.3 μatm ( 平均297.3 ± 56.2 μatm, n = 1113 )，低值 ( 109.1 – 299.4μatm, n = 313 ) 主要出現在低溫低鹽高營養鹽的長江沖淡水 ( Salinity (S) < 31.5 ) 之擴散面積 ( 6.4×1010 m2, 佔研究區域之34.9% )，主要受生物作用影響。而高值(338.4 – 366.1 μatm , n = 95 )主要出現在高溫高鹽低營養的黑潮水 ( S > 34)；至2003年8月觀測期間表水fCO2範圍介於 235.0 - 467.2 μatm ( 平均373.3 ± 33.2 μatm , n = 1624 )，低值 ( 235.0 – 322.7 μatm , n = 75 ) 依然出現在長江沖淡水(面積1.2×1010 m2，佔研究區域的4.4%)，其表水fCO2略高於6月之數值；而高溫高鹽的黑潮水，其表水fCO2略高於6月之數值範圍介於 369.8 – 409.4 μatm ( n = 462 )。2003年6月三峽大壩進行第一階段蓄水，短短兩個月的時間，各水型有明顯的消長變化，2003年8月 ( 1.2×1010 m2 ) 長江沖淡水擴散面積較6月 (6.4×1010 m2 ) 減少81.2 %，造成生物作用降低，水溫提高，因此2003年6月至8月二氧化碳海氣交換通量由 -1.9 ± 1.6 mole C m-2 y-1轉變為0.2 ± 0.9，明顯的由一個大氣二氧化碳的「匯」(Sink) 轉為「源」(Source)。 夏季東海表水fCO2的分布主要與側向傳輸的混合機制有關係，此外各水型之間的主控因素亦不相同，黑潮水與陸棚混合水 ( 31.5 < S < 34 ) 主要受控於溫度變化，長江沖淡水主要受控於生物作用，沿岸水 ( Depth < 50m ) 主要受控於垂直混合作用。研究顯示溫度、鹽度與葉綠素a代表各控制機制的參數，並將各水型分離，於水深 > 50m的陸棚海域上進而建立了四組多變數回歸經驗模式 ( r2 > 0.83 )，其中模式1 （fCO2 = -776.209 + (73.830 × T) - (1.169 × T2) - (26.018 × Chla) + (1.356 × Chla2)）可利用遙測之溫度、葉綠素a的資料進行多變數回歸換算，模式結果顯示，在2003年夏季時東海的代表性通量為 -0.8 ± 1.0 mole C m-2 y-1，為大氣二氧化碳的「匯」( sink )，若無生物作用時則增加為 0.2 ± 0.9 mole C m-2 y-1，此亦使得表水二氧化碳分壓增加9.3% ( 研究區域內之平均由332.1 μatm轉為363.0 μatm )。而2003年夏季( 6月1日至8月31日 ) 之東海 ( 9 × 1011 m2 ) 約可吸收 8×10-3 Gt C，若無生物之作用則將釋放 2.5×10-3 Gt C。

The distribution and air-sea exchange fluxes of CO2 with its controlling factors had been investigated in surface waters of the East China Sea (ECS) by an automated underway CO2 system in summer (June 19th to June 26th and August 13th to 23rd) 2003. There were no significant differences in atmospheric CO2 between June (365.9 1.7 atm) and August 2003 (359.8 3.6 atm). Sea surface fCO2 distribution, for instance, in June 2003 showed significant spatial and temporal variation, ranging from 109.1 to 435.3 atm with an average of 297.3 56.2 atm (n1113). Low fCO2 levels (109.1 – 299.4 atm, n = 313), mainly due to bio-uptake effect, were found in low-temperature, low-salinity and nutrient-rich Changjiang Diluted waters (CDW, S<31.5) off Mainland China which coverage area of 6.41010 m2 accounted for 34.9% of the study area. High fCO2 values (338.4-366.1 atm, n=95) occurred mainly in high-temperature, high-salinity and nutrient-less Kuroshio waters (S>34). In August 2003, the sea surface CO2 ranged from 235.0 to 467.2 uatm (average 373.3 33.2 atm, n 1624). Low fCO2 (235.0 – 322.7 atm, n = 75) observed at the CDW (1.21010 m2, 4.4% of the study area) in August was slightly higher than that in June. Additionally, the CO2 levels in the Kuroshio waters (369.8–409.4 μatm, n = 462) are slightly higher in August than June. The results showed significant changes in water masses occurred in the ECS shelf area between June and August, 2003. The CDW coverage area was additionally decreased by 81.2% from June to August, resulting in less biological effect and a SST increase. Consequently, the air-sea CO2 exchange of the entire ECS shelf area from a sink (-1.9 ± 1.6 mole C m-2 y-1) to a source (0.2 ± 0.9 mole C m-2 y-1) happened within 2 months just right after the filling of the Three Gorge Dam (TGD) in June 2003. The relationship results showed the summer CO2 distribution was associated with mixing mechanisms via lateral transport in the ECS. Each type of water masses had their own major factors for governing CO2. CO2 changes in Kuroshio and shelf mixed waters were, for instance, mainly controlled by temperature effect, in CDW governed by biological activity and in costal waters (depth <50m) influenced by vertical mixing. Four multi-variable regression relationships were established (r2 > 0.83) according to temperature, salinity and chlorophyll a as major controlling variables. However, Model I relation, expressed as a polynomial of two parameters, was applied to the areas of water depth above 50 m ( Depth > 50m，fCO2 = -776.209 + (73.830 × T) - (1.169 × T2) - (26.018 × Chla) + (1.356 × Chla2)) by using remote sensing data of temperature and chlorophyll a. The model results showed that the ECS shelf area ( 9 × 1011 m2) in summer (June-August) 2003 acted as an atmospheric CO2 sink with an average air-to-sea flux of -0.8 ± 1.0 mole C m-2 y-1 which could absorb approximately 8×10-3 Gt C. If biological activities were turned off, the sea surface CO2 increased by 9.32% from an average of 332 μatm to 363 μatm and turned a sink to a source with a flux of 0.2 ± 0.9 mole C m-2 y-1 i.e. 2.5×10-3 Gt C released into the atmosphere from the ECS.