2023-01-012024-05-13https://scholars.lib.ntu.edu.tw/handle/123456789/652792古氣候在不同pCO2濃度下的溫度變化，包括末次冰盛期 (LGM) 和上 新世中期暖期 (MPWP)，可提升對全球地表熱傳輸機制的 了解，亦可作為預測未來的氣候模型之基準。 重建地球過去的地表溫度可以 從保存在海洋沉積物中的地球化學指標得出，而這些指標通常被稱 為代用指標，由於其並不是直接測量海水溫度，因此它們的解釋時 常並非直觀的，同時代用指標在重建時耗費時間和資源故多重代用 指標數據稀少，至今仍無法解決一個根本問題:不同溫度代用指標 可以交換使用來推斷過去的氣候嗎?為此，該計畫旨在太平洋地區 驗證代用指標的訊號來源並建立多重代用指標的溫度數據庫，主要 針對黑潮、南海和赤道太平洋地區。第 1 年的計劃以獲取沉積物岩 心樣本、招聘人員和生成多重代用指標(UK`37、TEX86、多種屬 Mg/Ca)的數據。在本計畫的前 10 個月內，通過與日本、德國和澳 大利亞的同事合作，我們為 LGM 數據庫(25-0 ka)獲得 8 根岩心 ，為 MPWP 數據庫(3.4-2.4 Ma)獲得 3 根岩心，結合之前的 MOST 項目(2018-2021 年)中生成的多重代用指標溫度數據集，新 添加的數據將我們的多重代用指標溫度數據庫中的沉積物岩心數量 增加到 14 個，使其成為我們所知的最大數據庫。在前 10 個月內 已經產生了一些初步結果，在LGM 數據庫的顯著結果包括(1) 日 本外海(31oN 和 34oN)的兩個 UK`37 記錄顯示相似的溫度演變 ，最低溫發生在 ~16 ka。(2)與上述日本外海記錄相反，台灣外 海沿黑潮和琉球海流路徑的 UK`37、TEX86 、Mg/Ca 記錄顯示 LGM 期間有最低溫。(3)在 3 個代用指標和模擬所推斷的LGM冷卻程度 中，TEX86始終強於UK`37、Mg/Ca 和由高解析度區域海洋模型 sbPOM，而源自 TEX86 的異常強烈的 LGM 冷卻可能是由於全新世和 LGM 期間古菌棲息地深度的變化。在驗證代用指標的訊號來源中顯 著結果有:(1)沉積物收集器 UK`37 和 TEX86 時間序列中缺乏季 節性波動，而居住於表水的有孔蟲Trilobatus sacculifer Mg/Ca 時間序列中的季節性週期減弱但有明顯 3oC的變化。(2) 台灣西 南近海水柱中的烯烴和古細菌四醚低到無法測量(每個樣品過濾 15 L 海水)。(3)與貝氏統計和普通最小二乘 TEX86 校準相比，基 於機器學習的 Optimal GDGT 校準產生更大的變化幅度。該計畫目 前得到 2名博士生和 6 名碩士生的支持，他們的工作為驗證代用指 標的訊號來源和 LGM 數據庫做出了貢獻。我們將在項目的第二年和 第三年申請博士後研究員幫助。來自首爾國立大學的候選人將為 MPWP 數據庫做出貢獻。 Paleoclimate data from key geological intervals, such as the Last Glacial Maximum (LGM) and Mid Pliocene Warm Period (MPWP), may improve mechanistic understanding of heat transport at Earth’s surface and serve as a benchmark for state-of-the-art climate models used to project future climate change. Past surface temperature on Earth can be derived from geochemical indicators embedded in marine sediments typically known as proxies. Proxies are not based on direct measurement of seawater temperature, thus their interpretation is not always straightforward. Due to the time- and resource-consuming nature of proxy reconstruction, multiproxy records are yet scarce, precluding a systematic effort to address a fundamental question: Can temperature proxies be used interchangeably to infer past climate? To this end, this project aims to ground-truth proxies and generate multiproxy-based temperature database in the Pacific Ocean, with emphasis on the Kuroshio Current, South China Sea, and Equatorial Pacific. The planned activities for the 1st year are to obtain sediment core samples, recruit personnel, and generate multiproxy (UK’37, TEX86, multi-species Mg/Ca) data. Within the first 10 months of the project, we obtained 8 cores for the LGM database (25–0 ka) and 3 for the MPWP database (3.4–2.4 Ma), via collaboration with colleagues in Japan, Germany and Australia. Combined with the multiproxy temperature datasets generated in my previous MOST project (2018–2021), the newly added LGM cores increase the number of sediment cores in our multiproxy temperature database to 14, rendering it the largest database to our knowledge. Notable preliminary results for the LGM database include (1) two UK’37 records off Japan (31o and 34 oN) show similar temperature evolution, with lowest temperatures occurring at ~16 ka. (2) In contrast to the aforementioned records off Japan, UK’37, TEX86, and Mg/Ca records off Taiwan along the path of Kuroshio and Ryukyu Current show the lowest temperatures during the LGM instead. (3) Here, of the 3 proxies, TEX86- inferred LGM cooling is always stronger than that based on UK’37 and Mg/Ca, as well as than the LGM cooling estimates derived from high-resolution regional ocean model sbPOM. The unusually strong TEX86-derived LGM cooling might be due to changing archaeal habitat depths during the Holocene and LGM. Notable preliminary results for proxy ground-truthing include (1) the lack of seasonal fluctuations in the sediment trap UK’37 and TEX86 time series, in contrast to an attenuated but evident 3oC of seasonal cycle in the Mg/Ca time series of surface-dwelling foraminifera Trilobatus sacculifer, (2) low to undetectable alkenones and archaeal tetraethers in the water column offshore southwest Taiwan (15 L seawater filtered per sample), (3) machine-learning-based Optimal GDGT calibration results in larger magnitude of change compared to Bayesian and ordinary least square TEX86 calibrations. The project is now supported by 2 PhD students and 6 MSc students, whose work contributes to proxy ground-truthing and the LGM database. We will apply for postdoctoral funding for the 2nd and 3rd year of the project. The candidate, who is from Seoul National University, will contribute to the MPWP database.古海洋學;末次盛冰期;上新世中期暖期;海表溫度;古氣候代用 指標;黑潮; 南海;赤道太平洋;UK’37; TEX86;鎂/鈣;Paleoceanography; Last Glacial Maximum; Mid Pliocene Warm Period; Sea surface temperature; Paleoclimate proxies; Kuroshio; South China Sea; Equatorial Pacific; UK’37; TEX86; Mg/Ca