| 摘要: | 由於2020-2021年沒有颱風登陸，台灣面臨水資源問題，導致2021年出現春季乾旱，對需要大量供水的工農業部門造成影響。遙測技術的進步催生了各種乾旱指數來獲取乾旱期間的地表反應。地表水可用性和溫度 (SWAT) 指數結合了歸一化差異潛熱指數 (NDLI)、歸一化差異植被指數 (NDVI) 和地表溫度 (LST)，使用歐式距離估算地表乾燥度。本研究建議將NDLI、NDVI 和LST 的測定值(R²) 與氣象乾旱指數（包括帕爾莫嚴重度指數 (PDSI) 和標準化降雨蒸發散指數 (SPEI）的相關性相結合來調整SWAT。此調整方法的結果可用於利用調整後的 SWAT 及其與聖嬰現象ENSO 的相關性來評估乾旱事件的時空特徵，使用柯本氣候分類法來描述地帶特徵；根據SWAT，使用MODIS來反演地表乾燥狀況，並使用TerraClimate來反演氣象乾旱，研究時間範圍為2002年至2021年的1月至5月，重點在於春季乾旱。結果顯示，SPEI-6 最適合調整SWAT，改善與植被水分脅迫和氣象乾旱的相關性，顯示台灣地表條件更容易受到較長時期乾旱的影響。台灣大部分地區經歷潮濕或正常的表面乾燥水平，偶爾有輕度、中度和重度乾燥，主要發生在著重農業的西部和南部地區；減少的趨勢發生在人類住區附近，而增加的趨勢發生在偏遠的森林地區，這表示人為影響發生在可到達的地區。主要影響聖嬰現象的相關方向是地形和地理特徵，不是氣候帶，而這影響最初的降水模式。;Due to a lack of typhoon landfalls in 2020-2012, Taiwan has faced water resource problems, leading to a spring drought in 2021 that impacted both industrial and agricultural sectors, which require extensive water supplies. Advances in remote sensing technology have led to various drought indices to capture surface responses during droughts. The Surface Water Availability and Temperature (SWAT) index combines the Normalized Difference Latent Heat Index (NDLI), Normalized Difference Vegetation Index (NDVI), and Land Surface Temperature (LST) to estimate surface dryness using Euclidean distance. This study proposes a modified SWAT by incorporating the coefficient of determination (R²) of NDLI, NDVI, and LST with meteorological drought indices, including the Palmer Drought Severity Index (PDSI) and Standardized Precipitation-Evapotranspiration Index (SPEI). The findings from this adjustment method are used to assess the spatial and temporal characteristics of drought events using the modified SWAT and its correlation with El Niño-Southern Oscillation (ENSO). The Köppen-Geiger climate classification is used to describe zonal characteristics. MODIS is used to retrieve surface dryness conditions based on SWAT, and TerraClimate data is used to retrieve meteorological drought. The study period ranges from January to May, from 2002 to 2021, focusing on spring droughts. The results reveal that SPEI-6 is the most suitable for SWAT weighting, improving correlations with vegetation water stress and meteorological drought, suggesting that Taiwan′s surface conditions are more influenced by longer periods of drought. Taiwan mostly experienced wet or normal surface dryness levels, with occasional mild, moderate, and severe dryness, predominantly in the western and southern regions linked to agricultural practices. Decreasing trends occurred near human settlements, while increasing trends were observed in remote forest areas, suggesting anthropogenic influences in accessible regions. Topographical and geographical features, rather than climate zones, primarily influence correlation directions with ENSO phases, which initially affect precipitation patterns. |
