以機器學習預測海溫及熱帶氣旋特徵對於珊瑚白化之影響 – 以澎湖南方四島為例

以作者查詢圖書館館藏

、以作者查詢臺灣博碩士

、以作者查詢全國書目

、勘誤回報

、線上人數：10

、訪客IP：3.133.92.215

姓名

馮馨柔(Shan-Non Feng) 查詢紙本館藏

畢業系所

土木工程學系

論文名稱

以機器學習預測海溫及熱帶氣旋特徵對於珊瑚白化之影響 – 以澎湖南方四島為例
(Applying Machine Learning to Predict the SST Variation and Tropical Cyclone Patterns on Influence of Coral Bleaching – Using South Penghu Marine National Park as an example)

相關論文

★ 結合資料探勘方法建立屏東平原含水層水文地質參數推估模式	★ 探討颱風特性於農損及坡地災害遙測影像辨識之研究
★ 不同時空降雨型態對於地下水補注量之探討—以屏東平原為例	★ 以訊號分析資料探勘方法探討PM2.5污染傳播時空特徵及相應之天氣條件
★ 運用訊號分析方法於地下水資源旱災韌性與風險評估	★ 探討都市熱島效應對臺北地區午後雷雨及地下水之影響
★ 水文地質條件不確定性下的地下水時空變化模擬	★ 建立台灣北部交通與氣象因子對於空氣污染影響之機器學習模型
★ 以深度學習方法建立地下水位預警之風險評估模型	★ 探討臺灣地震活動特徵與環境變數相關性分析
★ 以機器學習方法建立巨觀尺度降雨氣候水資源推估模式

檔案

[Endnote RIS 格式]

[Bibtex 格式]

[相關文章]

[文章引用]

[完整記錄]

[館藏目錄]

[檢視]

[下載]

本電子論文使用權限為同意立即開放。
已達開放權限電子全文僅授權使用者為學術研究之目的，進行個人非營利性質之檢索、閱讀、列印。
請遵守中華民國著作權法之相關規定，切勿任意重製、散佈、改作、轉貼、播送，以免觸法。

摘要(中)

2020 年，位於澎湖南方海域的澎湖南方四島發生嚴重的珊瑚白化現象且可能造成大量珊瑚死亡，珊瑚為海洋中重要的生態庫，豐富海中的生物多樣性，但珊瑚可以存活的溫度範圍狹窄，暖化造成的海溫(Sea Surface Temperature，SST) 上升影響不容小覷，珊瑚的復原時間也長達 10 年，若極端氣候持續的發生，珊瑚將會更難生存，故能準確預測未來之海溫變化極為重要。
除了海溫外，澎湖附近海域之珊瑚白化指數 DHW (Degree Heating Weeks) 也在 2020 年達到史上最高，且近 15 年珊瑚白化風險透過分析結果發現有著明顯的上升。本研究以深度學習方法，長短期記憶模型(Long Short-Term Memory, LSTM)結合卷積神經網路(Convolutional Neural Network, CNN)來預測海溫，並可以利用預測之海溫反推算 DHW，海溫歷史資料之準確度高達 99.22%，對於預測未來海溫及 DHW 可以提供良好貢獻。
澎湖在 2020 年也並未直接受到颱風的侵襲，少了颱風攪動海水，冷卻海溫的效果，連續的高溫也是其中原因導致大面積的珊瑚白化，而颱風的強度、風速及海洋混和層厚度都會造成不同程度的海溫冷卻，本研究分析了過去颱風事件得到的不同因子對於海溫冷卻之相關性及貢獻程度外，也建立了機器學習模型隨機森林分類器(Random Forest Classifier)預測並分類不同因子之颱風事件是否造成顯著海溫冷卻效應，透過歷史資料預測也有 92.9%之準確率，其中以經驗正交函數(Empirical Orthogonal Function, EOF)對於海溫的分組提供最顯著之海溫冷卻因子。
除了分析數據資料外，本研究也到澎湖南方四島進行了 2 次現地調查，在 2021 年澎湖南方四島之珊瑚並未受到明顯珊瑚白化之影響。本研究建立海溫及 DHW 預測模型與颱風冷卻分類模型，希望未來可以提供相關機構預測之海溫及珊瑚白化資訊，為澎湖南方四島之珊瑚保育盡一份心力。

摘要(英)

In 2020, the South Penghu Marine National Park had experienced sever coral bleaching and may cause coral deaths. It’s an important creature that enrich the biodiversity in the ocean. However, the SST that corals can survive is narrow, and the recovery time is about 10 years. The impact of extreme weather can’t be underestimate, and it is critical to accurately predict SST variations.
In addition to SST, DHW near Penghu achieved the highest level in 2020, the risk of coral bleaching has increased significantly in the past. In this study, the deep learning method, LSTM and CNN, is applied to predict SST, and is already reached 99.22% accuracy.
Penghu was not directly attacked by any typhoon in 2020, so lack of typhoon induced SST cooling effect, may be a reason that large area of coral bleaching. The intensity, wind speed, and the thickness of the ocean mixed layer would cause different degrees of SST cooling. A machine learning model, Random Forest, is established to classify whether this typhoon would cause significant cooling effect. The accuracy of the model is 92.9%, and the EOF analysis provides the most obvious cooling factor.
In addition to analyzing the data, this study also conducted twice site survey in Penghu Marine National Park in 2021, and corals were not affected by obvious coral bleaching. This research establishes SST, DHW prediction model and a typhoon induced SST cooling model, hope to provide forecast information for relevant institutions and contribute to the coral conservation of Penghu Marine National Park.

關鍵字(中)

★ 珊瑚白化
★ 海溫
★ 深度學習
★ 颱風特徵
★ DHW珊瑚白化閾值

關鍵字(英)

★ Coral bleaching
★ Sea surface temperature
★ Deep learning
★ Typhoon characteristics
★ Degree Heating Weeks

論文目次

摘要................................................i
Abstract ..........................................ii
致謝................................................iii
目錄................................................iv
表目錄..............................................vii 圖目錄.................................................ix 第一章
緒論............................................... 1
1.1 研究背景與動機................................... 1
1.2 研究問題與目的................................... 3
1.3 論文結構........................................ 5
第二章文獻回顧...................................... 6
2.1 珊瑚白化之相關研究................................ 7
2.2 颱風對於海溫造成之冷卻效應機制...................... 11
2.3 聖嬰-南方震盪現象 ................................ 14
2.4 深度學習模型應用於海溫預測.......................... 17
第三章研究方法....................................... 20
3.1 研究架構......................................... 20
3.2 研究區域概述...................................... 23
3.3 資料蒐集與描述.................................... 25
3.3.1 大氣因子資料.................................... 26
3.3.2 日本氣象廳，Best Track 颱風資料 ..................28
3.3.3 颱風資料庫...................................... 30
3.3.4 海溫 SST........................................31
3.4 小波訊號分析...................................... 35
3.5颱風類型指數 TTI ...................................37
3.6經驗正交函數 EOF ...................................39
3.7機器學習於海溫及颱風冷卻效應與否之判斷預測.............. 43
3.7.1 機器學習分類器.................................. 44
3.7.2 深度學習網路架構................................. 48
3.8 生態調查計畫...................................... 52
第四章結果分析與討論.................................. 57
4.1澎湖南方四島周邊海溫分析............................. 57
4.1.1 30 年海溫趨勢分析.................................57
4.1.2 小波訊號分析法................................... 59
4.1.3 ConvLSTM 模型預測海溫及反推算 DHW..................61
4.2 氣候因子分析....................................... 65
4.2.1 相關性分析....................................... 65
4.2.2 小波訊號分析法.................................... 67
4.2.3 多變量之海溫預測分析............................... 71
4.3 DHW 與其他參數之相關性分析............................72 4.3.1 DHW 之空間分布圖..................................72
4.3.2 小波訊號分析法................................... 75
4.4颱風事件分析........................................ 78
4.4.1 依中心距離分類.................................... 82
4.4.2 依路徑分類....................................... 84
4.4.3 颱風冷卻效應機制.................................. 86
4.4.4 依強度分類....................................... 89
4.4.5 EOF 分析.........................................90
4.4.6 TTI 分類 ........................................94
4.4.7 颱風對海溫冷卻效應預測模型.......................... 98
第五章結論與建議....................................... 107
5.1 結論.............................................. 107
5.2 討論.............................................. 108
5.3 建議.............................................. 110
5.4 貢獻.............................................. 111
參考文獻............................................... 112

參考文獻

[1] Barber, R. T. and F. P. Chávez (1986). "Ocean variability in relation to living resources during the 1982–83 El Niño." Nature 319(6051): 279-285.
[2] Breiman, L. (2001). "Random forests." Machine learning 45(1): 5-32.
[3] Breiman, L., et al. (1984). "Cart." Classification and Regression Trees;
Wadsworth and Brooks/Cole: Monterey, CA, USA.
[4] Brown, B. (1997). "Coral bleaching: causes and consequences." Coral reefs
16(1): S129-S138.
[5] Ding, M., et al. (2019). "A gated recurrent unit neural networks based wind
speed error correction model for short-term wind power forecasting."
Neurocomputing 365: 54-61.
[6] Edward, J. P., et al. (2018). "Coral mortality in the Gulf of Mannar, southeastern
India, due to bleaching caused by elevated sea temperature in." Current Science
114(9): 1967.
[7] Glynn, P. W., et al. (2001). "Coral bleaching and mortality in Panama and
Ecuador during the 1997–1998 El Niño–Southern Oscillation event: spatial/temporal patterns and comparisons with the 1982–1983 event." Bulletin of Marine Science 69(1): 79-109.
[8] Goreau, T., et al. (2000). "Conservation of coral reefs after the 1998 global bleaching event." Conservation Biology 14(1): 5-15.
[9] Goreau, T. J. and R. L. Hayes (1994). "Coral bleaching and ocean" hot spots"." Ambio-Journal of Human Environment Research and Management 23(3): 176- 180.
[10] Hochreiter, S. and J. J. N. c. Schmidhuber (1997). "Long short-term memory."9(8): 1735-1780.
[11] Hoegh-Guldberg, O. (1999). "Climate change, coral bleaching and the future of
the world′s coral reefs." Marine and freshwater research 50(8): 839-866.
[12] Hwang, P. A., et al. (2003). "A note on analyzing nonlinear and nonstationary
ocean wave data." Applied Ocean Research 25(4): 187-193.
[13] Jaimes, B. and L. K. Shay (2009). "Mixed Layer Cooling in Mesoscale Oceanic Eddies during Hurricanes Katrina and Rita." Monthly Weather Review 137(12):
4188-4207.
[14] Jokiel, P. L. and S. Coles (1977). "Effects of temperature on the mortality and
growth of Hawaiian reef corals." Marine Biology 43(3): 201-208.
[15] LeCun, Y., et al. (1998). "Gradient-based learning applied to document
recognition." Proceedings of the IEEE 86(11): 2278-2324.
[16] Lin, Y.-C., et al. (2020). "Typhoon Type Index: A New Index for Understanding the Rain or Wind Characteristics of Typhoons and Its Application to Agricultural Losses and Crop Vulnerability." Journal of Applied Meteorology and
Climatology 59(5): 973-989.
[17] Nalley, D., et al. (2016). "Inter-annual to inter-decadal streamflow variability in
Quebec and Ontario in relation to dominant large-scale climate indices." Journal
of hydrology 536: 426-446.
[18] NOAA (2020, 7/22). "Daily Global 5 km Satellite Coral Bleaching Heat Stress
Alert Area." Retrieved 12/21, 2020, from
https://coralreefwatch.noaa.gov/product/5km/index_5km_baa_max_r07d.php.
[19] Peñaflor, E., et al. (2009). "Sea-surface temperature and thermal stress in the
Coral Triangle over the past two decades." Coral reefs 28(4): 841-850.
[20] Pichot, V., et al. (2001). "Wavelet transform of heart rate variability to assess autonomic nervous system activity does not predict arousal from general
anesthesia." Canadian Journal of Anesthesia 48(9): 859-863.
[21] Prasad, T. G. and P. J. Hogan (2007). "Upper-ocean response to Hurricane Ivan in a 1/25° nested Gulf of Mexico HYCOM." Journal of Geophysical Research:
Oceans 112(C4).
[22] Price, J. F. (1981). "Upper ocean response to a hurricane." Journal of Physical
Oceanography 11(2): 153-175.
[23] Qin, Z., et al. (2008). "Characterization of CO2 and water vapor fluxes in a
summer maize field with wavelet analysis." Ecological Informatics 3(6): 397-
409.
[24] Richman, M. B. (1986). "Rotation of principal components." Journal of
climatology 6(3): 293-335.
[25] Sanford, T. B., et al. (1987). "Ocean Response to a Hurricane. Part I:
Observations." Journal of Physical Oceanography 17(11): 2065-2083.
[26] Sharma, A. K., et al. (2018). Analysis on the occurrence of tropical cyclone in the South Pacific region using recurrent neural network with LSTM.
International Conference on Neural Information Processing, Springer.
[27] Timmermann, A., et al. (2018). "El Niño–southern oscillation complexity."
Nature 559(7715): 535-545.
[28] Wada, A., et al. (2014). "Typhoon-induced sea surface cooling during the 2011
and 2012 typhoon seasons: Observational evidence and numerical investigations of the sea surface cooling effect using typhoon simulations." Progress in Earth and Planetary Science 1(1): 1-25.
[29] Wang, J., et al. (2020). "Spatiotemporal characteristics of PM2. 5 concentration in the Yangtze River Delta urban agglomeration, China on the application of big data and wavelet analysis." Science of The Total Environment: 138134.
[30] Wang, X., et al. (2011). "Impact of barrier layer on typhoon-induced sea surface
cooling." Dynamics of Atmospheres and Oceans 52(3): 367-385.
[31] Xie, J., et al. (2019). "An adaptive scale sea surface temperature predicting
method based on deep learning with attention mechanism." 17(5): 740-744.
[32] Yogo, M. (2008). "Measuring business cycles: A wavelet analysis of economic
time series." Economics Letters 100(2): 208-212.
[33] Yu, X., et al. (2020). "A novel method for sea surface temperature prediction
based on deep learning." Mathematical Problems in Engineering 2020.
[34] Zhu, T. and D.-L. Zhang (2006). "The impact of the storm-induced SST cooling
on hurricane intensity." Advances in atmospheric sciences 23(1): 14-22.
[35] 張至維, et al. (2013). 澎湖南方四島海域生態熱點調查與潛點規劃, 海洋國家公園管理處.

指導教授

林遠見(Yuan-Chien Lin)

審核日期

2023-7-5

推文