Extreme floods can be caused by various combinations of hydrological and meteorological factors and river basin conditions that have not been observed for a long time. Long-term observational series permit estimation of both the frequency and variation of spring floods – the key issues of protection systems. Fortunately, Baltic States have a long-term record of hydrological data for the last 80 years. In this research, spring flood parameters (maximum discharge, height of maximum discharge and its timing) for the Baltic countries were assessed for four periods (1922–2008, 1941–2008, 1961–2008 and 1991–2008). In total, 70 hydrological data series of spring flood parameters were used. To detect trends in time series for these periods, the Mann–Kendall test and the nonparametric Sen's method for the magnitude of the trend were used. The index flood method was used to estimate the maximum discharge in ungauged catchments. The results showed that maximum discharges and heights of spring floods decreased over a longer period. Spring flood peaks took place on earlier dates. Only some significant trends of maximum discharges and their timing were found in the last time period (1991–2008). All these changes could be caused by the increasing ambient temperature and precipitation in the later decades.

The river discharge changes in three Baltic States and its relation to changes in the main climatic variables such as precipitation and air temperature were analyzed using observed data and methods of empirical statistical analysis. The study is important for the development of efficient water resource management systems and validation of climate change impact models. The application of the Mann-Kendall test reveals that a significant increasing trend in winter air temperature and precipitation was determined for all 3 investigated periods (1923–2003, 1941–2003 and 1961–2003). The same trend was found for the winter and annual discharge time series. No trend was observed for the spring, summer and autumn seasonal streamflow and summer low flow series for most of the Baltic region. In general the relation between the main meteorological and hydrological parameters and the tendency in river discharge trends is common for all of the Baltic States, and might be associated with the regional impacts of global climate change.

Hydromorphology of rivers assessed through direct measurements is a time-consuming and relatively expensive procedure. The rapid development of unmanned aerial vehicles and machine learning (ML) technologies enables the usage of aerial images to determine hydromorphological units (HMUs) automatically. The application of various direct and indirect data sources and their combinations for the determination of river HMUs from aerial images was the main aim of this research. Aerial images with and without the Sobel filter, a layer of boulders identified using Yolov5x6, and a layer of direct measurements of depth and streamflow velocity were used as data sources. Three ML models were constructed for the cases if one, two, or three data sources were used. The ML models for HMU segmentation were constructed of MobileNetV2 pre-trained on ImageNet data for the feature extraction part and U-net for the segmentation part. The stratified K-fold cross-validation with five folds was carried out to evaluate the performance of the model due to the limited dataset. The analysis of the ML results showed that the measured metrics of segmentation using direct measurements were close to the ones of the model trained only on the combination of boulder layer and aerial images with the Sobel filter. The obtained results demonstrated the potential of the applied approach for the determination of HMUs only from the aerial images, and provided a basis for further development to increase its accuracy.