查詢結果分析
來源資料
頁籤選單縮合
| 題 名 | 氣候變遷下臺灣地區氣象乾旱單變數及雙變數分析=Univariate and Bivariate Analyses of Meteorological Droughts in Taiwan under Climate Change |
|---|---|
| 作 者 | 劉奕呈; 蕭政宗; | 書刊名 | 農業工程學報 |
| 卷 期 | 71:2 2025.06[民114.06] |
| 頁 次 | 頁22-40 |
| 分類號 | 430.1637 |
| 關鍵詞 | 氣候變遷; 乾旱; AR6統計降尺度日雨量資料; 標準化降水指數; 關聯結構; Climate change; Drought; AR6 statistical downscaling daily precipitation data; Standardized precipitation index; Copula; |
| 語 文 | 中文(Chinese) |
| DOI | 10.29974/JTAE.202506_71(2).0002 |
| 中文摘要 | 近年來氣候因為受到全球暖化的影響而發生變化,乾旱也不例外,未來乾旱的發生頻率 及嚴重性如何隨著氣候變遷的影響而變化逐漸受到重視。本研究旨在以單變數及雙變數分析 的方法探討臺灣地區氣候變遷對乾旱特性的影響,本文採用臺灣地區 AR6 統計降尺度網格 日雨量資料來評估乾旱特性的變化,包括 24 個全球氣候模式所產製四種不同暖化情境 (SSP1-2.6、SSP2-4.5、SSP3-7.0 及 SSP5-8.5) 的雨量,而日雨量資料的時程包含基期 (1995 ~ 2014) 及四個未來時期 (2021 ~ 2040、2041 ~ 2060、2061 ~ 2080 及 2081 ~ 2100)。本研究 首先將各網格日雨量資料轉換為一個月的標準化降水指數 (SPI) 以定義乾旱事件,並由乾旱 事件擷取三種乾旱特性 (頻率、延時及嚴重性),未來時期乾旱特性相對於基期的變化即顯示 氣候變遷對乾旱特性的影響。各網格的乾旱特性以 24 個全球氣候模式的系集平均為代表, 單變數分析的項目包括各乾旱特性在臺灣地區的空間分布變化、各水資源分區及未來不同時 段乾旱特性的差異。雙變數分析先以關聯結構 (copula) 建立乾旱延時與嚴重性雙變數機率 分佈,本文以三種乾旱事件強度 (乾旱延時與嚴重性機率分別為 0.25、0.5 及 0.75) 計算其聯 集重現期與交集重現期,再進行重現期在臺灣地區空間分布、水資源分區及未來不同時段變 化的分析。單變數分析結果顯示未來各時段不同暖化情境的三種乾旱特性相對於基期變化率 之絕對值由大到小依序為頻率、嚴重性、及延時,而雙變數聯集與交集重現期分析結果顯示 在世紀中至世紀末溫室氣體排放情境由輕微至嚴重依序為 SSP1-2.6、SSP2-4.5、SSP5-8.5 及 SSP3-7.0,單變數及雙變數乾旱分析均顯示臺灣北部及東部的乾旱特性受情境變化的影響較 中部及南部為大。 |
| 英文摘要 | There is an increasing concern on impacts of climate change on droughts since frequent and severe droughts are observed recently. This study aims to investigate the impacts of climate change on characteristics of drought events in Taiwan using both univariate and bivariate analysis methods. A total of 24 global climate models (GCMs) are used to statistically downscale projected daily precipitation into 0.5°×0.5° gridded precipitation data in Taiwan. The projection precipitation data consists of four emission scenarios (SSP1-2.6, SSP2-4.5, SSP3-7.0, and SSP5-8.5) and five time periods (baseline period of 1995 ~ 2014, future periods of 2021 ~ 2040, 2041 ~ 2060, 2061 ~ 2080, and 2081 ~ 2100,). The first step in drought analyses of this study is to transform daily rainfall series into 1-month Standardized Precipitation Index (SPI) series. A drought event is defined as the continuous negative SPI periods with a threshold of SPI < −1. Three drought characteristics (frequency, duration, and severity) are abstracted from each drought event. The univariate drought analysis includes spatial distribution, comparison among four regions, and temporal evolution in future periods of these three drought characteristics. Copula-based bivariate probability distributions of drought duration and severity for drought events are constructed in bivariate drought analysis. Three specific drought events with probabilities of 0.25, 0.5, and 0.75 for drought duration and severity are considered in this study to compute union and intersection return periods. The results of univariate drought analysis reveals that the relative changes of three drought characteristics from large to small are frequency, severity, and duration. In addition, impacts of climate change on these three drought characteristics from low (mild) to high (severe) are SSP1-2.6, SSP2-4.5, SSP5-8.5, and SSP3-7.0 over the period from middle to the end of the 21st century. The results of bivariate return periods indicate that severe drought events may occur under high-concentration emission scenarios (SSP3-7.0 and SSP5-8.5) with close return period. Among four regions in Taiwan, the northern and eastern regions have slightly shorter return periods, which means that these two regions are vulnerable to climate change. |
本系統中英文摘要資訊取自各篇刊載內容。