透過您的圖書館登入
IP:44.201.96.213
  • 學位論文

臺灣中部阿里山與玉山山區熱環境之趨勢

Trends of the Mountain Thermal Environment of Alishan and Yushan in Central Taiwan

指導教授 : 關秉宗

摘要


全球暖化是近年來相當熱門的議題,因為一地之熱環境變動不但影響生物生存,也攸關整個生態系功能的維持,而高山地區對氣候暖化較平地為敏感,更增加該地區物候相關研究的急迫性。然而,臺灣迄今尚缺乏特定山區物種的長期物候觀察資料,對於山區熱環境長期變化亦不甚瞭解。有鑑於此,本研究欲深入探討臺灣高山地區之熱環境趨勢,因此選擇中央氣象局阿里山測站與玉山測站之長期氣溫資料作為分析數據,分別代表臺灣中海拔與高海拔山區。先計算兩地每年春、秋、冬三季之累積生長度─日(growing degree-days, GDD),再根據世界氣候對照標準,計算得出「累積GDD距平(anomaly)」、「日最高溫距平」、「日最低溫距平」及每年累積至熱量基準之「所需日數距平」等四組時間序列。考慮氣候資料的非線性及非平穩特質,本研究以後設性訊號分離方法─「總體經驗模態分解法(ensemble empirical mode decomposition)」萃取兩組資料之趨勢,而後以一階與二階差分估算趨勢變化之速度與加速度,相較傳統方法更能深入瞭解趨勢性質,並用「最大熵值重複置還取樣法(maximum entropy bootstrap)」產生重複樣本,建立趨勢及其速度、加速度曲線之信賴區間,以判斷趨勢變動之關鍵年份。 結果顯示,兩地在三季的累積GDD都有逐年上升的趨勢,即環境熱量越來越高,並以秋季上升幅度最大。其中,阿里山三季累積GDD皆逐年以等加速度增加,玉山三季之趨勢速度雖亦逐年增加,但加速度自1980年代後開始逐漸下降,並於2010年代降至0值附近。而根據日最高溫、日最低溫距平時間序列分析結果,發現造成兩地差異之主因為日最低溫趨勢不同。另外,所需日數趨勢分析結果和累積GDD趨勢互相呼應,只是變成逐年減少,意即環境熱量累積的速度越來越快。比較不同的是,玉山累積GDD開始顯著上升之年份較阿里山延遲約5至10年,所需日數則兩地開始減少的年份相差不大。

並列摘要


Global warming has become an important research focus because the changes of thermal environment will affect not only the lives of organisms, but ecosystem functions as well. Alpine areas are more vulnerable to climate warming than low land areas. Hence, there is a pressing need for more phenological research in the former areas. However, Taiwan lacks long-term phenological data and has little understanding of its mountain thermal environment. Therefore, this study aims to probe into the trends of thermal environment in mountainous areas in Taiwan. The long-term temperature data of Alishan and Yushan weather stations, representing mid- and high-elevation areas in Taiwan, respectively, of the Central Weather Bureau were used in this study. The study first calculated the accumulated growing degree-days (GDD) of two stations in spring, autumn, and winter every year. Then, we obtained four time series— “GDD anomaly”, “Daily maximum temperature anomaly”, “Daily minimum temperature anomaly”, and “days anomaly” based on the 1961–1990 average. Because climate data are typically nonlinear and non-stationary, this study used “ensemble empirical mode decomposition” to extract the trends of four datasets, 1st differencing and 2nd differencing to approximate velocity and acceleration of the extracted trends, and “maximum entropy bootstrap” to establish confidence intervals for the estimates. The results showed that both stations had an upward trend in GDD anomaly in the three seasons, which means increasing heat accumulation in both areas, and autumn had the fastest increasing trend. Furthermore, in Alishan, the velocities of GDD anomaly trends in three seasons kept increasing with constant accelerations; in Yushan, though the velocities also kept increasing, the decelerations had begun since the 1980s and reached zero around 2010s. By analyzing the daily maximum and minimum temperature anomaly data, this study found the divergence between the GDD of Alishan and Yushan was mainly due to the opposite trends in the daily minimum temperatures. The results of days anomaly showed a similar but downward trend, compared to the GDD anomaly data, which indicates that the speed of heat accumulation became faster in both areas. The main difference between the two stations is that the GDD anomaly of Alishan started to increase for 5 to 10 years before that of Yushan. However, days anomaly began to decrease at almost the same time in the two stations.

參考文獻


Ahmad, L., Kanth, R. H., Parvaze, S., & Mahdi, S. S. (2017). Growing Degree Days to Forecast Crop Stages. In Experimental Agrometeorology: A Practical Manual, 95-98. Springer, Cham. https://doi.org/10.1007/978-3-319-69185-5_14
Alexandrov, T. (2008). A method of trend extraction using singular spectrum analysis. RevStat, 7 (1), 1-22. https://doi.org/10.1057/978-1-137-40951-5_3
Bühlmann, P. (1998). Sieve bootstrap for smoothing in nonstationary time series. The Annals of Statistics, 26(1), 48-83. https://doi.org/10.1214/aos/1030563978
Bale, J. S., Masters, G. J., Hodkinson, I. D., Awmack, C., Bezemer, T. M., Brown, V. K., et al. (2002). Global change biology, 8(1), 1-16. https://doi.org/10.1046/j.1365-2486.2002.00451.x
Barbosa, S., Scotto, M., & Alonso, A. (2011). Summarising changes in air temperature over Central Europe by quantile regression and clustering. Natural Hazards and Earth System Sciences, 11(12), 3227-3233. https://doi.org/10.5194/nhess-11-3227-2011

延伸閱讀