Search

River-flow forecasts are important for the management and planning of water resources and their rational use. The present study, based on direct multistep-ahead forecasting with multiple time series specific to the XGBoost algorithm, estimates the long-term changes and forecast monthly flows of selected rivers in Ukraine. In a new, applied approach, a single multioutput model was proposed that forecasts over both short- and long-term horizons using grouped or hierarchical data series. Three forecast stages were considered: using train and test subsets, using a model with train-test data, and training with all data. The historical period included the measurements of the monthly flows, precipitation, and air temperature in the period 1961–2020. The forecast horizons of 12, 60, and 120 months into the future were selected for this dataset, i.e., December 2021, December 2025, and December 2030. The research was conducted for diverse hydrological systems: the Prut, a mountain river; the Styr, an upland river; and the Sula, a lowland river in relation to the variability and forecasts of precipitation and air temperature. The results of the analyses showed a varying degree of sensitivity among rivers to changes in precipitation and air temperature and different projections for future time horizons of 12, 60, and 120 months. For all studied rivers, variable dynamics of flow was observed in the years 1961–2020, yet with a clearly marked decrease in monthly flows during in the final, 2010–2020 decade. The last decade of low flows on the Prut and Styr rivers was preceded by their noticeable increase in the earlier decade (2000–2010). In the case of the Sula River, a continuous decrease in monthly flows has been observed since the end of the 1990s, with a global minimum in the decade 2010–2020. Two patterns were obtained in the forecasts: a decrease in flow for the rivers Prut (6%) and the Styr (12–14%), accompanied by a decrease in precipitation and an increase in air temperature until 2030, and for the Sula River, an increase in flow (16–23%), with a slight increase in precipitation and an increase in air temperature. The predicted changes in the flows of the Prut, the Styr, and the Sula rivers correspond to forecasts in other regions of Ukraine and Europe. The performance of the models over a variety of available datasets over time was assessed and hyperparameters, which minimize the forecast error over the relevant forecast horizons, were selected. The obtained RMSE parameter values indicate high variability in hydrological and meteorological data in the catchment areas and not very good fit of retrospective data regardless of the selected horizon length. The advantages of this model, which was used in the work for forecasting monthly river flows in Ukraine, include modelling multiple time series simultaneously with a single model, the simplicity of the modelling, potentially more-robust results because of pooling data across time series, and solving the "cold start" problem when few data points were available for a given time series. The model, because of its universality, can be used in forecasting hydrological and meteorological parameters in other catchments, irrespective of their geographic location. [ABSTRACT FROM AUTHOR]

The paper investigates the distribution properties of measurement series of river water temperatures for the lowland River Note'c and its tributaries (Western Poland), as well as air temperatures at different data time resolution levels (1987-2013). The aspect of distribution normality was examined in quantile plots, the series' stationarity was assessed with an augmented Dickey-Fuller test, while autocorrelation was studied using an Autoregressive Integrated Moving Average (ARIMA) model. It was demonstrated that distributions of river water and air temperature series at different levels of analyses are generally close to normal but also display a certain skewness. Both daily temperature measurement series are stationary series. The periodic component accounts for about 93% (water temperature) and 77% (air temperature) of the daily variability of the variable, while the random factor equals 6-7% and 22%, respectively. The Autoregressive Integrated Moving Average (ARIMA) model confirmed a clear annual seasonality in temperature distribution and indicated the long memory of the autoregressive process AR (2-4). The temperature prediction performed on the basis of a 4th-order Fourier series is consistent with the course of historical data. In the multiannual period 1987-2013, particularly high maximum temperatures were recorded for the Upper Notec in the summer half-years (28.4 °C); these are related to anthropogenic factors and increase the threat to the existence of cyprinids and salmonids. The thermal anomalies identified in the River Notec clearly point to the necessity of intensifying the monitoring of its waters. [ABSTRACT FROM AUTHOR]

Forecasting ice phenomena in river systems is of great importance because these phenomena are a fundamental part of the hydrological regime. Due to the stochasticity of ice phenomena, their prediction is a difficult process, especially when data sets are sparse or incomplete. In this study, two machine learning models—Multilayer Perceptron Neural Network (MLPNN) and Extreme Gradient Boosting (XGBoost)—were developed to predict ice phenomena in the Warta River in Poland in a temperate climate zone. Observational data from eight river gauges during the period 1983–2013 were used. The performance of the model was evaluated using four model fit measures. The results showed that the choice of input variables influenced the accuracy of the developed models. The most important predictors were the nature of phenomena on the day before an observation, as well as water and air temperatures; river flow and water level were less important for predicting the formation of ice phenomena. The modeling results showed that both MLPNN and XGBoost provided promising results for the prediction of ice phenomena. The research results of the present study could also be useful for predicting ice phenomena in other regions. [ABSTRACT FROM AUTHOR]

The temperature of river water (TRW) is an important factor in river ecosystem predictions. This study aims to compare two different types of numerical model for predicting daily TRW in the Warta River basin in Poland. The implemented models were of the stochastic type—Autoregressive (AR), Moving Average (MA), Autoregressive Moving Average (ARMA) and Autoregressive Integrated Moving Average (ARIMA)—and the artificial intelligence (AI) type—Adaptive Neuro Fuzzy Inference System (ANFIS), Radial Basis Function (RBF) and Group Method of Data Handling (GMDH). The ANFIS and RBF models had the most fitted outputs and the AR, ARMA and ARIMA patterns were the most accurate ones. The results showed that both of the model types can significantly present suitable predictions. The stochastic models have somewhat less error with respect to both the highest and lowest TRW deciles than the AIs and were found to be better for prediction studies, with the GMDH complex model in some cases reaching Root Mean Square Error (RMSE) = 0.619 °C and Nash-Sutcliff coefficient (NS) = 0.992, while the AR(2) simple linear model with just two inputs was partially able to achieve better results (RMSE = 0.606 °C and NS = 0.994). Due to these promising outcomes, it is suggested that this work be extended to other catchment areas to extend and generalize the results. [ABSTRACT FROM AUTHOR]

The identification of changes in the ice phenomena (IP) in rivers is a significant element of analyses of hydrological regime features, of the risk of occurrence of ice jam floods, and of the ecological effects of river icing (RI). The research here conducted aimed to estimate the temporal and spatial changes in the IP in a lowland river in the temperate climate (the Noteć River, Poland, Central Europe), depending on air temperature (TA) and water temperature (TW) during the multi-annual period of 1987–2013. Analyses were performed of IP change trends in three RI phases: freezing, when there appears stranded ice (SI), frazil ice (FI), or stranded ice with frazil ice (SI–FI); the phase of stable ice cover (IC) and floating ice (FoI); and the phase of stranded ice with floating ice (SI–FoI), frazil ice with floating ice (FI–FoI), and ice jams (IJs). Estimation of changes in IP in connection with TA and TW made use of the regression model for count data with a negative binomial distribution and of the zero-inflated negative binomial model. The analysis of the multi-annual change tendency of TA and TW utilized a non-parametric Mann–Kendall test for detecting monotonic trends with Yue–Pilon correction (MK–YP). Between two and seven types of IP were registered at individual water gauges, while differences were simultaneously demonstrated in their change trends over the researched period. The use of the Vuong test confirmed the greater effectiveness of estimates for the zero-inflated model than for the temporal trend model, thanks to which an increase in the probability of occurrence of the SI phenomenon in the immediate future was determined; this, together with FI, was found to be the most frequently occurring IP in rivers in the temperate climate. The models confirmed that TA is the best estimator for the evaluation of trends of the occurrence of IC. It was shown that the predictive strength of models increases when thermal conditions are taken into consideration, but it is not always statistically significant. In all probability, this points to the impact of local factors (changes in bed and valley morphology and anthropogenic pressure) that are active regardless of thermal conditions and modify the features of the thermal-ice regime of rivers at specific spatial locations. The results of research confirm the effectiveness of compilating a few models for the estimation of the dependence of IP trends on air and water temperature in a river. [ABSTRACT FROM AUTHOR]

The study determined water temperature trends of rivers in Poland in the period 1971–2015, and also their spatial and temporal patterns. The analysis covered daily water temperature of 53 rivers recorded at 94 water gauge stations and air temperature at 43 meteorological stations. Average monthly, annual, seasonal and maximum annual tendencies of temperature change were calculated using the Mann–Kendall (M–K) test. Regional patterns of water temperature change were determined on the basis of Ward's hierarchical grouping for 16 correlation coefficients of average annual water temperature in successive 30-year sub-periods of the multi-annual period of 1971–2015. Moreover, regularities in monthly temperature trends in the annual cycle were identified using 12 monthly values obtained from the M–K Z test. The majority of average annual air and water temperature series demonstrate statistically significant positive trends. In three seasons: spring, summer and autumn, upward tendencies of temperature were detected at 70%–90% of the investigated water gauges. In 82% of the analysed rivers, similarity to the tendencies of change of monthly air temperature was concluded, with the climatic factor being recognised as of decisive importance for the changes in water thermal characteristics of the majority of rivers in Poland. In the winter months, positive trends of temperature were considerably weaker and in general statistically insignificant. On a regional scale, rivers with a quasi-natural thermal regime experienced temperature increases from April to November. In the other cases, different directions of change in river water temperature (RWT) were attributed to various forms of human impact. It was also found that for the majority of rivers the average annual water temperature in the analysed 30-year sub-periods displayed upward trends, statistically significant or close to the significance threshold. Stronger trends were observed in the periods after 1980, while a different nature of water temperature change was detected only in a couple of mountainous rivers or rivers transformed by human impact. In the beginning of the analysed period (1971–2015), the average annual water temperature of these rivers displayed positive and statistically significant trends, while after 1980 the trends were negative. The detected regularities and spatial patterns of water temperature change in rivers with a quasi-natural regime revealed a strong influence of climate on the modification of their thermal regime features. Rivers characterised by a clearly different nature of temperature change, both in terms of the direction of the tendencies observed and their statistical significance, were distinguished by alterations of water thermal characteristics caused by human activity. The results obtained may be useful in optimising the management of aquatic ecosystems, for which water temperature is a significant indicator of the ongoing environmental changes. [ABSTRACT FROM AUTHOR]

The objective of the paper consisted in determining the effect of macroscale types of NAO, AO, EA, EAWR, SCAND, and AMO atmospheric circulation on changes in water temperature in Polish rivers. The study has made use of a broad body of hydrometeorological materials covering daily water temperature values for 96 water gauge stations located on 53 rivers and air temperature values for 43 meteorological stations. Percentage shares of positive and negative coefficients of correlation of annual, seasonal, and monthly circulation type indices with air and river water temperature were determined, demonstrating the character of teleconnection. Determinations were made of water temperature deviations in positive and negative phases of the analyzed indices from average values from the years 1971–2015, and their statistical significance ascertained. Research has shown that relations between the temperature of river waters in Poland and macroscale circulation types are not strong, however they are noticeable, sometimes even statistically significant, and both temporally and spatially diverse. NAO, AO, EA, and AMO indices are characterized by a generally positive correlation with temperature, whereas SCAND and EWAR indices are characterized by a negative correlation. Research showed a varying impact of types of atmospheric circulation, with their effectiveness increasing in the winter season. The strongest impact on temperature was observed for the positive and negative NAO and AO phases, when deviations of water temperature from average values are correspondingly higher (up to 1.0 °C) and lower (by a maximum of 1.5 °C), and also for the positive and negative SCAND phases, when water temperature are correspondingly lower (by a maximum of 0.8 °C) and higher (by 1.2 °C) than average values. The strongest impact on water temperature in summer, mainly in July, was observed for AMO. The results point to the complexity of processes determining the thermal regime of rivers and to the possibility of additional factors—both regional and local—exerting an influence on their temporal and spatial variability. [ABSTRACT FROM AUTHOR]

An analysis was conducted of the activity of individual homogeneous Hydrological Response Units (HRUs) and their impact on the components of the spatially distributed water balance based on the example of two urbanised catchments of the city of Poznań (Western Poland). Water balance was developed using the WetSpass model and GIS spatial data, based on hydrometeorological data from the reference period of 1961–2000 including projected land usage changes and precipitation changes expected by 2025 in the city. The catchments were parameterised with reference to land usage, soil permeability, terrain declivities and the level of groundwater waters in summer and winter. The dependence between HRUs and their impact on components of the water balance was determined. Water balance forecasts have shown two-way changes in the components of approximately 12% of the catchments. A significant increase of surface runoff (an increase of 20–30 mm/HRU) at the expense of effective infiltration reduction (by 15–20 mm/HRU) was determined for arable land intended for development. An increase of infiltration and evapotranspiration at the expense of the surface runoff reduction is forecast for areas designed for urban afforestation. The tendency of increase of atmospheric precipitation within the city until 2025 was indicated by changes in the water balance components. Changes in the landscape resulting from urban expansion may lead to detrimental hydrological effects: accumulation of surface runoffs and occurrence of local flash flooding, as confirmed by the simulations carried out using the WetSpass model. The results may contribute to a more accurate understanding of the impact of urban landscape modification patterns on the water balance at the regional and local scale. [ABSTRACT FROM AUTHOR]

The article presents the results of an analysis of the impact of cumulative negative air temperature degree-days (CDD) on formation of ice cover on the typical lowland river in the moderate climate zone (River Noteć, western Poland). During the period 1987–2013 the occurrence of a series of 306 consecutive days of negative air temperatures was identified in relation to atmospheric circulation in the winter seasons, with a duration of 1–53 days. The longest CDD series occurred in 2010, while low air temperatures were particularly frequent in the year 1996 (−600 °C·d). The analysis showed that this phenomenon was most identifiable in December. Research made it possible to distinguish two types of circulation, which were conducive to the cumulation of CDD. Using the logistic regression model it was determined that the probability of ice cover occurrence increased most rapidly in the lower course (−70 °C·d), while the weakest interdependence was observed along the other sections. On average, a one-degree-day increase of CDD was connected with an increase in the probability of a permanent ice cover appearing of approximately 1.2–5.7%. The results obtained are of particular significance for identifying the thermal and circulatory factors determining the appearance of ice cover on the rivers of the Polish Lowlands. [ABSTRACT FROM AUTHOR]