This study aims to develop an advanced deep learning model, Hydro-Informer, for accurate water level and flood predictions, emphasizing extreme event forecasting. Utilizing a comprehensive dataset from the Slovak Hydrometeorological Institute SHMI (2008–2020), which includes precipitation, water level, and discharge data, the model was trained using a ladder technique with a custom loss function to enhance focus on extreme values. The architecture integrates Recurrent and Convolutional Neural Networks (RNN, CNN), and Multi-Head Attention layers. Hydro-Informer achieved significant performance, with a Coefficient of Determination (R²) of 0.88, effectively predicting extreme water levels 12 h in advance in a river environment free from human regulation and structures. The model’s strong performance in identifying extreme events highlights its potential for enhancing flood management and disaster preparedness. By integrating with diverse data sources, the model can be used to develop a well-functioning warning system to mitigate flood impacts. This work proposes a novel architecture suitable for locations without water regulation structures.

The water sources located on the inland islands on the left side of the Danube River as well as water sources located on the right bank of the Danube River in the Pečniansky forest have an irreplaceable function in terms of supplying the population in Bratislava with drinking water. There are 80 wells in these mentioned areas operated by the Bratislava Water Company. In order to maintain the actual functions of the water resources, it is therefore necessary to assess the impact of the proposed construction of the multi-purposed structure ‘Hydraulic structure Bratislava — Pečniansky forest’ on these water sources.

Presented paper deals with the hydraulic assessment of groundwater flow in the area affected by the realization of the hydraulic gate on the Klátov branch and in the adjacent territory of a dike, which is located on the right-side of Little Danube. This hydraulic assessment is part of the project of the Slovak Water Management Enterprise, which also aims to increase the height and seal the dike on the right-side of the Little Danube. Generally, the project is divided into three phases (Phase I, II and III) to implement different technical measures to protect the area from flooding. The assumption for the execution of the technical measures of the mentioned three project phases is a continuous flood protection of part of the Žitný ostrov area around the Little Danube and the Klátov River branch in the reach from Kolárovo to Jahodná town. Therefore, a 3D mathematical model was created to simulate groundwater flow by changing boundary conditions of surface water flow during flood periods.

The Bratislava region in Slovakia aims to improve its transport infrastructure by connecting the airport with the railway network. As part of the Trans-European Network for Transport project (TEN-T), an underground railway line is proposed to be constructed on both sides of the Danube River, connecting the airport in Bratislava to the Petržalka region on the river’s right side. However, underground construction is likely to have an impact on the groundwater flow regime. This construction, which will be built below the ground surface, should be built by excavating from above under the protection of sealing walls to prevent significant changes to the groundwater level regime (GWL). Therefore, a numerical model based on the finite element method (FEM) was established to evaluate the effect of the planned underground construction on the GWL, and technical measures were introduced to mitigate any potential impacts. The results of the model revealed possibilities for controlling the groundwater level in the aquifer affected by the railway structure during and after the construction.

The research presented in this paper focuses on the optimization and modelling of an innovative groundwater extraction system and addresses the challenge of managing the increased groundwater level in Nitra Industrial Park, where the Land Rover Jaguar car manufacturer is located. The primary objective of this research was to determine the optimal pumping rate of the system, ensuring the maintenance of the groundwater level at the required elevation without depleting the surrounding water resources. To achieve this goal, a numerical groundwater modelling approach was utilized, employing the TRIWACO simulation package and the finite element method (FEM). The study involved various simulations under steady and unsteady flow conditions, using hydrological, geological and hydrogeological data, including the results of three proposed pumping tests conducted by INGEO, Ltd. Company, as well as measured groundwater level data from boreholes monitored by the Slovak Hydrometeorological Institute (SHMI). Implementing transient groundwater flow simulations considered the impact of a 1000-year flash flood wave (Q1000). This paper presents results of numerical modelling results, offering insights into maintaining the required groundwater elevation while considering key parameters and uncertainties for design purposes. Additionally, on-site optimization was performed to ensure efficient system operation and minimize associated costs. In conclusion, this research contributes to sustainable groundwater management in the Nitra Industrial Park, addressing the challenge of the increased groundwater level and providing valuable findings to optimize the extraction system’s performance.