This work investigates historical trends of meteorological drought in Taiwan by means of long-term precipitation records. Information on local climate change over the last century is also presented. Monthly and daily precipitation data for roughly 100 years, collected by 22 weather stations, were used as the study database. Meteorological droughts of different levels of severity are represented by the standardized precipitation index (SPI) at a three-monthly time scale. Additionally, change-point detection is used to identify meteorological drought trends in the SPI series. Results of the analysis indicate that the incidence of meteorological drought has decreased in northeastern Taiwan since around 1960, and increased in central and southern Taiwan. Long-term daily precipitation series show an increasing trend for dry days all over Taiwan. Finally, frequency analysis was performed to obtain further information on trends of return periods of drought characteristics. Résumé Cet article étudie les tendances historiques de la sécheresse météorologique à Taiwan sur la base de longues séries de précipitations. Les informations relatives au changement climatique au cours du siècle dernier sont également présentées. Les données de précipitations mensuelle et journalière sur environ 100 ans, collectées par 22 stations météorologiques, ont été utilisées comme base de l'étude. Les sécheresses météorologiques de niveaux de sévérité différents sont représentées selon l'indice standardisé de précipitation (ISP) à pas de temps trimestriel. De plus, la détection de rupture est utilisée pour identifier les tendances de sécheresse météorologique dans les séries d'ISP. Les résultats de l'analyse indiquent que l'incidence de la sécheresse météorologique a diminué dans le nord-est de Taiwan depuis 1960 environ, et augmenté dans le centre et le sud de Taiwan. Les longues séries de précipitation journalière présentent une tendance croissante pour les jours secs pour l'ensemble de Taiwan. ...

Probabilistic flood forecasting, which provides uncertain information in the forecasting of floods, is practical and informative for implementing flood-mitigation countermeasures. This study adopted various machine learning methods, including support vector regression (SVR), a fuzzy inference model (FIM), and the k-nearest neighbors (k-NN) method, to establish a probabilistic forecasting model. The probabilistic forecasting method is a combination of a deterministic forecast produced using SVR and a probability distribution of forecast errors determined by the FIM and k-NN method. This study proposed an FIM with a modified defuzzification scheme to transform the FIM’s output into a probability distribution, and k-NN was employed to refine the probability distribution. The probabilistic forecasting model was applied to forecast flash floods with lead times of 1–3 hours in Yilan River, Taiwan. Validation results revealed the deterministic forecasting to be accurate, and the probabilistic forecasting was promising in view of a forecasted hydrograph and quantitative assessment concerning the confidence level.