Like many other African countries, incidence of drought is increasing in Nigeria. In this work, spatiotemporal changes in droughts under different representative concentration pathway (RCP) scenarios were assessed; considering their greatest impacts on life and livelihoods in Nigeria, especially when droughts coincide with the growing seasons. Three entropy-based methods, namely symmetrical uncertainty, gain ratio, and entropy gain were used in a multi-criteria decision-making framework to select the best performing General Circulation Models (GCMs) for the projection of rainfall and temperature. Performance of four widely used bias correction methods was compared to identify a suitable method for correcting bias in GCM projections for the period 2010-2099. A machine learning technique was then used to generate a multi-model ensemble (MME) of the bias-corrected GCM projection for different RCP scenarios. The standardized precipitation evapotranspiration index (SPEI) was subsequently computed to estimate droughts from the MME mean of GCM projected rainfall and temperature to predict possible spatiotemporal changes in meteorological droughts. Finally, trends in the SPEI, temperature and rainfall, and return period of droughts for different growing seasons were estimated using a 50-year moving window, with a 10-year interval, to understand driving factors accountable for future changes in droughts. The analysis revealed that MRI-CGCM3, HadGEM2-ES, CSIRO-Mk3-6-0, and CESM1-CAM5 are the most appropriate GCMs for projecting rainfall and temperature, and the linear scaling (SCL) is the best method for correcting bias. The MME mean of bias-corrected GCM projections revealed an increase in rainfall in the south-south, southwest, and parts of the northwest whilst a decrease in the southeast, northeast, and parts of central Nigeria. In contrast, rise in temperature for entire country during most of the cropping seasons was projected. The results further indicated that increase in temperature would decrease the SPEI across Nigeria, which will make droughts more frequent in most of the country under all the RCPs. However, increase in drought frequency would be less for higher RCPs due to increase in rainfall.
Globally, many studies on machine learning (ML)-based flood susceptibility modeling have been carried out in recent years. While majority of those models produce reasonably accurate flood predictions, the outcomes are subject to uncertainty since flood susceptibility models (FSMs) may produce varying spatial predictions. However, there have not been many attempts to address these uncertainties because identifying spatial agreement in flood projections is a complex process. This study presents a framework for reducing spatial disagreement among four standalone and hybridized ML-based FSMs: random forest (RF), k-nearest neighbor (KNN), multilayer perceptron (MLP), and hybridized genetic algorithm-gaussian radial basis function-support vector regression (GA-RBF-SVR). Besides, an optimized model was developed combining the outcomes of those four models. The southwest coastal region of Bangladesh was selected as the case area. A comparable percentage of flood potential area (approximately 60% of the total land areas) was produced by all ML-based models. Despite achieving high prediction accuracy, spatial discrepancy in the model outcomes was observed, with pixel-wise correlation coefficients across different models ranging from 0.62 to 0.91. The optimized model exhibited high prediction accuracy and improved spatial agreement by reducing the number of classification errors. The framework presented in this study might aid in the formulation of risk-based development plans and enhancement of current early warning systems.