A linear regression machine learning model to estimate the reference evapotranspiration based on temperature data for South Korea is developed in this study. FAO56 Penman–Monteith (FAO56 P–M) reference evapotranspiration calculated with meteorological data (1981–2021) obtained from sixty-two meteorological stations nationwide is used as the label. All study datasets provide daily, monthly, or annual values based on the average temperature, daily temperature difference, and extraterrestrial radiation. Multiple linear regression (MLR) and polynomial regression (PR) are applied as machine learning algorithms, and twelve models are tested using the training data. The results of the performance evaluation of the period from 2017 to 2021 show that the polynomial regression algorithm that learns the amount of extraterrestrial radiation achieves the best performance (the minimum root-mean-square errors of 0.72 mm/day, 11.3 mm/month, and 40.5 mm/year for daily, monthly, and annual scale, respectively). Compared to temperature-based empirical equations, such as Hargreaves, Blaney–Criddle, and Thornthwaite, the model trained using the polynomial regression algorithm achieves the highest coefficient of determination and lowest error with the reference evapotranspiration of the FAO56 Penman–Monteith equation when using all meteorological data. Thus, the proposed method is more effective than the empirical equations under the condition of insufficient meteorological data when estimating reference evapotranspiration.

This study aims to assess the spatiotemporal characteristics of agricultural droughts in Bangladesh during 1981–2015 using the Effective Drought Index (EDI). Monthly precipitation data for 36 years (1980–2015) obtained from 27 metrological stations, were used in this study. The EDI performance was evaluated for four sub-regions over the country through comparisons with historical drought records identified by regional analysis. Analysis at a regional level showed that EDI could reasonably detect the drought years/events during the study period. The study also presented that the overall drought severity had increased during the past 35 years. The characteristics (severity and duration) of drought were also analyzed in terms of the spatiotemporal evolution of the frequency of drought events. It was found that the western and central regions of the country are comparatively more vulnerable to drought. Moreover, the southwestern region is more prone to extreme drought, whereas the central region is more prone to severe droughts. Besides, the central region was more prone to extra-long-term droughts, while the coastal areas in the southwestern as well as in the central and north-western regions were more prone to long-term droughts. The frequency of droughts in all categories significantly increased during the last quinquennial period (2011 to 2015). The seasonal analysis showed that the north-western areas were prone to extreme droughts during the Kharif (wet) and Rabi (dry) seasons. The central and northern regions were affected by recurring severe droughts in all cropping seasons. Further, the most significant increasing trend of the drought-affected area was observed within the central region, especially during the pre-monsoon (March–May) season. The results of this study can aid policymakers in the development of drought mitigation strategies in the future.

The impacts of climate change on precipitation and drought characteristics over Bangladesh were examined by using the daily precipitation outputs from 29 bias-corrected general circulation models (GCMs) under the representative concentration pathway (RCP) 4.5 and 8.5 scenarios. A precipitation-based drought estimator, namely, the Effective Drought Index (EDI), was applied to quantify the characteristics of drought events in terms of the severity and duration. The changes in drought characteristics were assessed for the beginning (2010–2039), middle (2040–2069), and end of this century (2070–2099) relative to the 1976–2005 baseline. The GCMs were limited in regard to forecasting the occurrence of future extreme droughts. Overall, the findings showed that the annual precipitation will increase in the 21st century over Bangladesh; the increasing rate was comparatively higher under the RCP8.5 scenario. The highest increase in rainfall is expected to happen over the drought-prone northern region. The general trends of drought frequency, duration, and intensity are likely to decrease in the 21st century over Bangladesh under both RCP scenarios, except for the maximum drought intensity during the beginning of the century, which is projected to increase over the country. The extreme and medium-term drought events did not show any significant changes in the future under both scenarios except for the medium-term droughts, which decreased by 55% compared to the base period during the 2070s under RCP8.5. However, extreme drought days will likely increase in most of the cropping seasons for the different future periods under both scenarios. The spatial distribution of changes in drought characteristics indicates that the drought-vulnerable areas are expected to shift from the northwestern region to the central and the southern region in the future under both scenarios due to the effects of climate change.

Storage rate forecasting for the agricultural reservoir is helpful for preemptive responses to disasters such as agricultural drought and planning so as to maintain a stable agricultural water supply. In this study, SVM, RF, and ANN machine learning algorithms were tested to forecast the monthly storage rate of agricultural reservoirs. The storage rate observed over 30 years (1991–2022) was set as a label, and nine datasets for a one- to three-month storage rate forecast were constructed using precipitation and evapotranspiration as features. In all, 70% of the total data was used for training and validation, and the remaining 30% was used as a test. The one-month storage rate forecasting showed that all SVM, RF, and ANN algorithms were highly reliable, with R² values ≥ 0.8. As a result of the storage rate forecast for two and three months, the ANN and SVM algorithms showed relatively reasonable explanatory power with an average R² of 0.64 to 0.69, but the RF algorithm showed a large generalization error. The results of comparing the learning time showed that the learning speed was the fastest in the order of SVM, RF, and ANN algorithms in all of the one to three months. Overall, the learning performance of SVM and ANN algorithms was better than RF. The SVM algorithm is the most credible, with the lowest error rates and the shortest training time. The results of this study are expected to provide the scientific information necessary for the decision-making regarding on-site water managers, which is expected to be possible through the connection with weather forecast data.