利用基隆和長潭里潮位站資料估算平均海水位

臺灣採用正高高程系統，係利用1957年至1991年基隆潮位站資料計算求得之平均海水位（mean sea level），此系統命名為2001臺灣高程基準（TaiWan Vertical Datum 2001，TWVD 2001），並以水準原點K999作為起算點、K998為副點。然而，民國99年（西元2010年）副點K998滅失，且周圍環境已無法滿足地質穩固之需求。民國103年（西元2014年）內政部同意將水準原點遷至新建置之水準原點K997，同時間在附近建置長潭里潮位站和全球導航衛星系統（Global Navigation Satellite System, GNSS）連續觀測站，該潮位站之平均海水位可作為日後水準零點，以利後續建置一正高高程系統。本研究評估基隆和長潭里潮位站資料品質後，使用1991年後近30年基隆潮位站水位資料，計算不同時期平均海水位，分析與TWVD2001水準零點差異之變化；亦考慮長潭里潮位資料之節點潮、逆氣壓、基準偏移影響，並計算平均海水位。此外，本研究亦利用歷年水準檢測成果和GNSS連續觀測量計算長潭里潮位站之地表垂直變動量，分析其周圍地質是否穩固。成果顯示四種方法（算術平均法、調和分析法、六參數擬合及多參數線性擬合）計算之2004-2022年（時間中點2013.5）基隆平均海水位平均值高於TWVD2001水準零點約67.3 mm，故現今基隆平均海水位與TWVD2001水準零點已有顯著差異變化。而計算之2015-2022年（時間中點2019.15）長潭里平均海水位高於TWVD2001水準零點為121.0 mm，和利用基隆潮位站資料推估時間中點2019.15之平均海水位（82.0 mm）有39.0 mm差異，其差值可能是長潭里潮位站資料涵蓋時間較短（2015-2022）且資料品質較差（前期資料未進行氣象署品管處理）、兩潮位站海水面地形不同、基隆潮位站參考基準未釐清（如基隆潮位站資料無法呈現水準檢測之每年5 mm地表下沉量）等所致。最後，2015~2023長潭里GNSS解呈現地表高程變化速率為0.3±0.1 mm/yr，與歷年水準檢測成果相符，顯示長潭里潮位站周圍地表穩定平穩。因此建議未來持續監測基隆和長潭里平均海水位高。

The vertical datum adopted by Taiwan is orthometric height, which is based on the mean sea level derived from Keelung tide gauge records covering from 1957 to 1991. This datum is named the TaiWan Vertical Datum 2001(TWVD 2001) and the original benchmark is K999, with the associate datum benchmark designed as K998. However, K998 was lost in 2010, and the surrounding environment no longer meets the requirements for geological stability. In 2014, the Ministry of the Interior consented to move the original benchmark to the newly established K997. Concurrently, a tide gauge station and a continuously operating Global Navigation Satellite System (GNSS) station were instituted at Zhangtanli, proximate to K997. The mean sea level derived from this tide gauge can serve as a future reference (zero level) for establishing a subsequent orthometric height system. The study evaluates the data quality of Keelung and Zhangtanli tide gauge records. The Keelung tide gauge records covering the past 30 years after 1991 were used to calculate the mean sea level at different time spans, and then we analyze the differences compared to the TWVD2001 original benchmark. Additionally, this study also considers the effects of nodal tide, inverse barometer, and datum offset on the Zhangtanli tide gauge data, and calculate the mean sea level. Furthermore, the study also uses levelling results and GNSS continuous observation data to calculate the vertical motion of the Zhangtanli tide gauge and analyzes whether the surrounding geological environment is stable. The results indicate that the mean sea surfaces in Keelung, as calculated by four different algorithms (arithmetic mean, harmonic analysis, six-parameter linear regression, multiple-variable linear regression) using tide gauge records from 2004 to 2022 (midpoint in time being 2013.5), are approximately 67.3 mm above the TWVD2001 zero level. The computed mean sea surface at Zhangtanli tide gauge from 2015 to 2022 (midpoint in time being 2019.15) is 121.0 mm higher than the zero level of TWVD2001. Compared to the estimated mean sea level of 82.0 mm at the Keelung tide station for the same midpoint in time, a discrepancy of approximately 39.0 mm is observed. The discrepancy may be due to several factors such as the shorter time span (2015-2022) and lower quality of the Zhangtanli tide gauge records (the preliminary data were not subjected to quality control by the Central Weather Administration), variations in sea surface topographies between the two tide gauge stations, and unclarified reference datum for the Keelung tide gauge station (for example, the Keelung tide gauge records do not reflect the annual ground subsidence of−5 mm measured by levelling). Lastly, the GNSS solutions at Zhangtanli from 2015 to 2023 show a vertical land motion rate of 0.3±0.1 mm/yr, which is identical to the levelling results, suggesting a consistent and stable surface elevation surrounding the Zhangtanli tide gauge station. Therefore, it is recommended that the mean sea surfaces at Keelung and Zhangtanli tide gauges be continuously monitored in the future.