透過您的圖書館登入
IP:3.80.164.96
  • 學位論文

台灣區域未來降雨推估的統計降尺度穩定性研究

The study of statistical downscaling stationarity for future precipitation projection over Taiwan

指導教授 : 陳正達
若您是本文的作者,可授權文章由華藝線上圖書館中協助推廣。

摘要


目前CMIP5(Coupled Model Intercomparison Project Phase 5)所使用的氣候模式其空間解析度對於區域性的地區或國家(如台灣)在評估未來氣候變遷的影響時仍有所不足,此時為了克服氣候模式其解析度較低的限制,降尺度方法的運用便成為研究區域氣候的必要手段。 統計降尺度方法因其簡便且計算的需求相對較少,因此此方法已廣泛應用於全球各個區域的氣候研究上且行之有年。而統計降尺度方法在應用於未來氣候推估的降尺度時,其前提皆是假設過去(或現今)資料所建立的統計關係到未來時仍是穩定的;然而,近年來氣候變遷日趨嚴重,已有人開始質疑此無法驗證(在無未來的觀測資料情況下)的前提假設在未來是否仍成立。因此本研究採用"理想模式"("perfect model")此實驗架構利用高解析度的模式資料(動力降尺度資料)代替原本降尺度中所使用的觀測資料(因模式資料有模擬未來的部分),以驗證誤差修正氣候特徵法(Bias corrected Climate Imprint,簡稱BCCI)、誤差修正建構類比法(Bias corrected Constructed Analogues,簡稱BCCA)和誤差修正建構類比兼分位映射法(BCCA with quantile mapping reordering,簡稱BCCAQ)此三種統計降尺度方法在應用於未來的降尺度時能否遵守上述的前提假設,並比較不同統計降尺度方法其結果在現在和未來時期的表現,以及在這兩個時期表現的穩定性。 研究結果顯示,BCCA此降尺度方法降尺度後的日降雨結果在強度上皆有低估的情況,BCCI和BCCAQ的結果在強度和極端降雨指標(r1mm、rx1day、rx5day)的表現上則與原始高解析度的模式資料較相近;至於統計穩定性的評估是以平均絕對誤差的比值(未來/現在)是否大於1而定,大於1即表示統計降尺度方法在應用於未來的降尺度時,其誤差會比應用於現在時期要來的大,此也代表違反了上述的前提假設。而本研究所驗證的三種統計降尺度方法(BCCI、BCCA、BCCAQ)其比值皆大於1,其中BCCA最大,其次為BCCAQ,BCCI則最小,此也表示BCCA的統計穩定性表現較差,BCCI則表現較佳。

並列摘要


The spatial resolution of the climate models currently used by CMIP5 is still insufficient for regional regions or countries (such as Taiwan) in assessing the impact of future climate change. At this time, in order to overcome the limitation of the lower resolution of the climate models, the application of the downscaling method has become a necessary approach for studying regional climate. The statistical downscaling method is simple and requires relatively fewer computing resources. Therefore, this method has been widely used in climate research in various regions of the world for many years. The statistical downscaling method when applied to the downscaling of future climate projections, assumes that the statistical relationships established by past (or present) data will remain stable in the future. However, in recent years, climate change has become increasingly serious, and some people have begun to question whether this premise assumption that can’t be verified (in the absence of future observations) is still valid in the future. Therefore, this study uses the "perfect model" experimental design, this experimental design used high-resolution model data (dynamic downscaling data) instead of observation data in downscaling (because the model can simulate the future part) to verify whether bias corrected climate imprint (BCCI), bias corrected constructed analogues (BCCA), and bias corrected constructed analogues with quantile mapping reordering (BCCAQ) this three statistical downscaling methods can comply with the above that premise assumption when applyed to the future projections downscaling and compare the performance of different statistical downscaling methods in the present and future periods and the stability performance in these two periods. The results of this study show that the BCCA method underestimates in intensity on its daily precipitation results. By contrast, the daily precipitation results of BCCI and BCCAQ are similar to the original high-resolution model data in terms of intensity and extreme precipitation indices. As for the evaluation of statistical stability, it depends on whether the ratio of the mean absolute error (future divided by present) is greater than 1. If the ratio greater than 1 indicates that when the statistical downscaling method is applied to future period, its error will be greater than that applied in the present period, which also represents a violation of the above premise assumption. The three statistical downscaling methods (BCCI, BCCA, BCCAQ) validated by this study all have ratios greater than 1, among which BCCA is the largest, followed by BCCAQ, and BCCI is the smallest. This also means that BCCA performed poorly on statistical stability, and BCCI performed better.

並列關鍵字

none

參考文獻


戴俐卉、洪景山、莊秉潔、蔡徵霖與倪佩貞,2008:WRF模式臺灣地區土地利用類型之更新與個案研究。大氣科學,36,43-62。
Barnston, A.G., M.H. Glantz, and Y. He, 1999: Predictive Skill of Statistical and Dynamical Climate Models in SST Forecasts during the 1997–98 El Niño Episode and the 1998 La Niña Onset. Bulletin of the American Meteorological Society,80,217–244,https://doi.org/10.1175/1520-0477(1999)080<0217:PSOSAD>2.0.CO;2
Bukovsky, M. S., and D. J. Karoly, 2009: Precipitation Simulations Using WRF as a Nested Regional Climate Model. Journal of Applied Meteorology and Climatology, 48, 2152–2159.
Cannon, A.J., S.R. Sobie, and Murdock T.Q., 2015: Bias Correction of GCM Precipitation by Quantile Mapping: How Well Do Methods Preserve Changes in Quantiles and Extremes?. Journal of Climate, 28, 6938–6959, https://doi.org/10.1175/JCLI-D-14-00754.1
Ciais, P., C. Sabine, G. Bala, L. Bopp, V. Brovkin, J. Canadell, A. Chhabra, R. DeFries, J. Galloway, M. Heimann, C. Jones, C. Le Quéré, R.B. Myneni, S. Piao and P. Thornton, 2013: Carbon and Other Biogeochemical Cycles. In Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change [Stocker, T.F., D. Qin, G.-K. Plattner, M. Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley (eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA.

延伸閱讀