Recent studies have shown the potential of transformer-based neural networks in increasing prediction capacity. However, classical transformers present several problems such as computational time complexity and high memory requirements, which make Long Sequence Time-Series Forecasting (LSTF) challenging. The contribution to the prediction of time series of flood events using deep learning techniques is examined, with a particular focus on evaluating the performance of the Informer model (a particular implementation of transformer architecture), which attempts to address the previous issues. The predictive capabilities of the Informer model are explored and compared to statistical methods, stochastic models and traditional deep neural networks. The accuracy, efficiency as well as the limits of the approaches are demonstrated via numerical benchmarks relating to real river streamflow applications. Using daily flow data from the River Test in England as the main case study, we conduct a rigorous evaluation of the Informer efficacy in capturing the complex temporal dependencies inherent in streamflow time series. The analysis is extended to encompass diverse time series datasets from various locations (>100) in the United Kingdom, providing insights into the generalizability of the Informer. The results highlight the superiority of the Informer model over established forecasting methods, especially regarding the LSTF problem. For a forecast horizon of 168 days, the Informer model achieves an NSE of 0.8 and maintains a MAPE below 10%, while the second-best model (LSTM) only achieves −0.63 and 25%, respectively. Furthermore, it is observed that the dependence structure of time series, as expressed by the climacogram, affects the performance of the Informer network.

In the context of implementing the European Flood Directive in Greece, a large set of rainfall data was compiled with the principal aim of constructing rainfall intensity–timescale–return period relationships for the entire country. This set included ground rainfall data as well as non-conventional data from reanalyses and satellites. Given the European declaration of climate emergency, along with the establishment of a ministry of climate crisis in Greece, this dataset was also investigated from a climatic perspective using the longest of the data records to assess whether or not they support the climate crisis doctrine. Monte Carlo simulations, along with stationary Hurst–Kolmogorov (HK) stochastic dynamics, were also employed to compare data with theoretical expectations. Rainfall extremes are proven to conform with the statistical expectations under stationarity. The only notable climatic events found are the clustering (reflecting HK dynamics) of water abundance in the 1960s and dry years around 1990, followed by a recovery from drought conditions in recent years.

Machine learning techniques have been increasingly used in flood management worldwide to enhance the effectiveness of traditional methods for flood susceptibility mapping. Although these models have achieved higher accuracy than traditional ones, their application has not yet reached full maturity. We focus on applying machine learning models to create flood susceptibility maps (FSMs) for Thessaly, Greece, a flood-prone region with extreme flood events recorded in recent years. This study utilizes 13 explanatory variables derived from topographical, hydrological, hydraulic, environmental and infrastructure data to train the models, using Storm Daniel—one of the most severe recent events in the region—as the primary reference for model training. The most significant of these variables were obtained from satellite data of the affected areas. Four machine learning algorithms were employed in the analysis, i.e., Logistic Regression (LR), Support Vector Machine (SVM), Random Forest (RF) and eXtreme Gradient Boosting (XGBoost). Accuracy evaluation revealed that tree-based models (RF, XGBoost) outperformed other classifiers. Specifically, the RF model achieved Area Under the Curve (AUC) values of 96.9%, followed by XGBoost, SVM and LR, with 96.8%, 94.0% and 90.7%, respectively. A flood susceptibility map corresponding to a 1000-year return period rainfall scenario at 24 h scale was developed, aiming to support long-term flood risk assessment and planning. The analysis revealed that approximately 20% of the basin is highly prone to flooding. The results demonstrate the potential of machine learning in providing accurate and practical flood risk information to enhance flood management and support decision making for disaster preparedness in the region.

Recent studies have shown the potential of Transformer-based neural networks in increasing prediction capacity. However, classical transformers present several problems such as computational time complexity and high memory requirements, which make Long Sequence Time-Series Forecasting (LSTF) challenging. The contribution to the prediction of time series of flood events using deep learning techniques is examined, with a particular focus on evaluating the performance of the Informer model (a particular implementation of transformer architecture), which attempts to address the previous issues. This study explores the predictive capabilities of the Informer model compared to statistical methods, stochastic models and traditional deep neural networks. The accuracy, efficiency as well as the limits of the approaches are demonstrated via numerical benchmarks relating to real river streamflow applications. Using daily flow data from the River Test in England as the main case study, we conduct a rigorous evaluation of the Informer efficacy in capturing the complex temporal dependencies inherent in flood time series. Among other things, the present work extends its analysis to encompass diverse time series datasets from various locations (>100) in the United Kingdom, providing insights into the generalizability of the Informer. The results highlight the significant superiority of the Informer model over established forecasting methods, especially regarding the LSTF problem. Furthermore, it is observed that the structure of time series, as expressed by climacogram, affects the performance of the Informer network.

Accurate precipitation estimations are crucial for hydrological modelling and water resource management, especially in geographically complex regions like Greece. Satellite-based products are valuable as they encompass extensive spatial coverage with high data density, but their accuracy is limited compared to ground measurements. To address this bias, we leverage machine learning and stochastic approaches. We present a hybrid machine learningstochastic framework that employs post-processing techniques to integrate satellite-derived precipitation data with ground-based gauge observations. We compare various independent machine learning algorithms and our approach for correcting satellite precipitation data over Greece. We use daily data for the period 2000-2021 from gauge observations and Integrated MultiSatellite Retrievals for Global Precipitation Measurement (IMERG). The framework enables stochastic outputs with uncertainty estimation of a model. The stochastic upgrade secured an 88.7% coverage rate for a 90% confidence level. We also propose a method capable of determining the output distribution, which can be used to generate reliable and accurate intensity–time scale–return period curves (also known as ombrian curves). For this purpose, we use the annual maximum daily rainfall over the same period, drawing on a denser network of stations for this component. The method’s output distribution achieved an NSE of 0.99 and a KGE of 0.96, compared to 0.78 and 0.65 for the satellite data. By integrating post-processing techniques within the learning pipeline, the proposed method improves accuracy over conventional merging approaches.

Machine learning techniques have been increasingly used in flood management worldwide to enhance the effectiveness of traditional methods for flood susceptibility mapping. Although these models have achieved higher accuracy than traditional ones, their application has not yet reached full maturity. We focus on applying machine learning models to create flood susceptibility maps (FSMs) for Thessaly, Greece, a flood-prone region with extreme flood events recorded in recent years. The study utilizes 13 explanatory variables derived from topographical, hydrological, hydraulic, environmental and infrastructure data to train the models, using the Daniel storm—one of the most severe recent events in the region—as the primary reference for model training. The most significant of these variables were obtained from satellite data of the affected areas. Four machine learning algorithms were employed in the analysis, i.e., Logistic Regression (LR), Support Vector Machine (SVM), Random Forest (RF), and eXtreme Gradient Boosting (XGBoost). Accuracy evaluation revealed that tree-based models (RF, XGBoost) performed better, achieving Area Under the Curve (AUC) values of 97%. A flood susceptibility map corresponding to a 1000-year return period rainfall scenario at 24 h scale was developed, aiming to support long-term flood risk assessment and planning. The results demonstrate the potential of machine learning in providing accurate and practical flood risk information to enhance flood management and support decision-making for disaster preparedness in the region.