Hydrological forecasting is a key factor for proper water resource management, as well as for monitoring and mitigating extreme events such as floods and droughts. However, due to climate changes observed in the last decades, conducting a proper evaluation of the meteorological-climatic parameters has become more complex. This has made hydrological forecasting a challenging task.
In literature, there are different approaches for hydrological modelling, e.g., physical-based models, which imitate the hydrological processes, and conceptual-based models, which use mathematical equations to represent the hydrological processes. However, physical-based models often require a large number of parameters with long time series to represent the hydrological cycle. At the same time, conceptual-based models require parameters that are not directly measurable, but must be evaluated based on observed data, which are not always available. Therefore, both approaches are complex to use for practical applications.
A further approach is represented by the Artificial Intelligence (AI) algorithms. These are increasingly implemented for the modelling of complex natural phenomena – inter alia, groundwater level, spring discharge, river flow rate, rainfall and evapotranspiration – since they do not require development of complex analytical relationships between inputs and target with a high computational speed.
Overall, this Research Topic aims to cover recent advances on the following topics:
• Development of innovative methods for hydrological forecasting, including individual and hybrid machine learning-based models;
• Spatio-temporal analysis of hydrological parameters aimed at investigating the complex relationships between the analyzed quantities and the geomorphological characteristics of the study areas; and,
• Development of models focused on the prediction of extreme hydrological events.
Hydrological forecasting is a key factor for proper water resource management, as well as for monitoring and mitigating extreme events such as floods and droughts. However, due to climate changes observed in the last decades, conducting a proper evaluation of the meteorological-climatic parameters has become more complex. This has made hydrological forecasting a challenging task.
In literature, there are different approaches for hydrological modelling, e.g., physical-based models, which imitate the hydrological processes, and conceptual-based models, which use mathematical equations to represent the hydrological processes. However, physical-based models often require a large number of parameters with long time series to represent the hydrological cycle. At the same time, conceptual-based models require parameters that are not directly measurable, but must be evaluated based on observed data, which are not always available. Therefore, both approaches are complex to use for practical applications.
A further approach is represented by the Artificial Intelligence (AI) algorithms. These are increasingly implemented for the modelling of complex natural phenomena – inter alia, groundwater level, spring discharge, river flow rate, rainfall and evapotranspiration – since they do not require development of complex analytical relationships between inputs and target with a high computational speed.
Overall, this Research Topic aims to cover recent advances on the following topics:
• Development of innovative methods for hydrological forecasting, including individual and hybrid machine learning-based models;
• Spatio-temporal analysis of hydrological parameters aimed at investigating the complex relationships between the analyzed quantities and the geomorphological characteristics of the study areas; and,
• Development of models focused on the prediction of extreme hydrological events.