Hydrological Forecasting Research Articles

Two Strategies for Avoiding Overfitting Long-Term Forecasting Models: Downsampling Predictor Fields and Shrinking Coefficients

Long-term hydrological forecasting based on sea surface temperature (SST) fields faces the large p and small n problem, i.e., too many potential predictors and a limited number of samples. Considering the selection of predictors will also enhance the complexity of models and lead to overfitting, in this study, two strategies are used for building forecasting models for long-term streamflow forecasting. The first strategy is to downsample the SST field and optimize its spatial resolution; the second is to shrink model coefficients based on L1 regularization. We build models based on the downsampled SST fields with different spatial resolutions. It is found that the model based on a proper spatial resolution always performs better than the model based on the raw SST field. This result suggests that it is better to treat the spatial resolution of the predictor field as a hyperparameter, which is similar to hyperparameters for controlling the complexity of many machine learning models. For applying the second strategy, L1 norm regularization models, including (1) least absolute selection and shrinkage operator (LASSO), (2) relaxed LASSO, and (3) two-step approach of LASSO and ordinary least squares regression (LASSO+OLS) are explored. We have found that the relaxed LASSO model always performs better than the ordinary LASSO, indicating that relaxed LASSO is a better shrinkage approach. Furthermore, the skills of the presented models are compared with the stepwise regression, and the lower skill of the stepwise regression based on SST fields with a high spatial resolution suggests that one should not select predictors based on fields with high resolutions considering the limited number of samples.

A New Method for Hour-by-Hour Bias Adjustment of Satellite Precipitation Estimates over Mainland China

Highly accurate near-real-time satellite precipitation estimates (SPEs) are important for hydrological forecasting and disaster warning. The near-real quantitative precipitation estimates (REGC) of the recently developed Chinese geostationary meteorological satellite Fengyun 4A (FY4A) have the advantage of high spatial and temporal resolution, but there are errors and uncertainties to some extent. In this paper, a self-adaptive ill-posed least squares scheme based on sequential processing (SISP) is proposed and practiced in mainland China to correct the real-time biases of REGC hour by hour. Specifically, the scheme adaptively acquires sample data by setting temporal and spatial windows and constructs an error-correction model based on the ill-posed least squares method from the perspectives of climate regions, topography, and rainfall intensity. The model adopts the sequential idea to update satellite precipitation data within time windows on an hour-by-hour basis and can correct the biases of real-time satellite precipitation data using dynamically changing parameters, fully taking into account the influence of precipitation spatial and temporal variability. Only short-term historical data are needed to accurately rate the parameters. The results show that the SISP algorithm can significantly reduce the biases of the original REGC, in which the values of relative bias (RB) in mainland China are reduced from 11.2% to 3.3%, and the values of root mean square error (RMSE) are also reduced by about 17%. The SISP algorithm has a better correction in humid and semi-humid regions than in arid and semi-arid regions and is effective in reducing the negative biases of precipitation in each climate region. In terms of rain intensity, the SISP algorithm can improve the overestimation of satellite precipitation estimates for low rain intensity (0.2–1 mm/h), but the correction for high rain intensity (>1 mm/h) needs further improvement. The error component analysis shows that the SISP algorithm can effectively correct the hit bias. This study serves as a valuable reference for real-time bias correction using short-term accumulated precipitation data.

A Stacking Ensemble Model of Various Machine Learning Models for Daily Runoff Forecasting

Improving the accuracy and stability of daily runoff prediction is crucial for effective water resource management and flood control. This study proposed a novel stacking ensemble learning model based on attention mechanism for the daily runoff prediction. The proposed model has a two-layer structure with the base model and the meta model. Three machine learning models, namely random forest (RF), adaptive boosting (AdaBoost), and extreme gradient boosting (XGB) are used as the base models. The attention mechanism is used as the meta model to integrate the output of the base model to obtain predictions. The proposed model is applied to predict the daily inflow to Fuchun River Reservoir in the Qiantang River basin. The results show that the proposed model outperforms the base models and other ensemble models in terms of prediction accuracy. Compared with the XGB and weighted averaging ensemble (WAE) models, the proposed model has a 10.22% and 8.54% increase in Nash–Sutcliffe efficiency (NSE), an 18.52% and 16.38% reduction in root mean square error (RMSE), a 28.17% and 18.66% reduction in mean absolute error (MAE), and a 4.54% and 4.19% increase in correlation coefficient (r). The proposed model significantly outperforms the base model and simple stacking model indicated by both the Friedman test and the Nemenyi test. Thus, the proposed model can produce reasonable and accurate prediction of the reservoir inflow, which is of great strategic significance and application value in formulating the rational allocation and optimal operation of water resources and improving the breadth and depth of hydrological forecasting integrated services.

Modelado hidrológico basado en el algoritmo KNN: una aplicación para el pronóstico de caudales diarios del río Ramis, Perú

The forecast of river stream flows is of significant importance for the development of early warning systems. Artificial intelligence algorithms have proven to be an effective tool in hydrological modeling data-driven, since they allow establishing relationships between input and output data of a watershed and thus make decisions data-driven. This article investigates the applicability of the k-nearest neighbor (KNN) algorithm for forecasting the mean daily flows of the Ramis river, at the Ramis hydrometric station. As input to the KNN machine learning algorithm, we used a data set of mean basin precipitation and mean daily flow from hydrometeorological stations with various lags. The performance of the KNN algorithm was quantitatively evaluated with hydrological ability metrics such as mean absolute percentage error (MAPE), anomaly correlation coefficient (ACC), Nash-Sutcliffe efficiency (NSE), Kling-Gupta efficiency (KGE') and the spectral angle (SA). The results for forecasting the flows of the Ramis river with the k-nearest neighbor machine learning algorithm reached high levels of reliability with flow lags of one and two days and precipitation with three days. The algorithm used is simple but robust to make short-term flow forecasts and can be integrated as an alternative to strengthen the daily hydrological forecast of the Ramis river.

Forecasting the Ensemble Hydrograph of the Reservoir Inflow based on Post-Processed TIGGE Precipitation Forecasts in a Coupled Atmospheric-Hydrological System

The quality of precipitation forecasting is critical for more accurate hydrological forecasts, especially flood forecasting. The use of numerical weather prediction (NWP) models has attracted much attention due to their impact on increasing the flood lead time. It is vital to post-process raw precipitation forecasts because of their significant bias when they feed hydrological models. In this research, ensemble precipitation forecasts (EPFs) of three NWP models (National Centers for Environmental Prediction (NCEP), United Kingdom Meteorological Office (UKMO) (Exeter, UK), and Korea Meteorological Administration (KMA) (SEOUL, REPUBLIC OF KOREA)) were investigated for six historical storms leading to heavy floods in the Dez basin, Iran. To post-process EPFs, the raw output of every single NWP model was corrected using regression models. Then, two proposed models, the Group Method of Data Handling (GMDH) deep learning model and the Weighted Average–Weighted Least Square Regression (WA-WLSR) model, were employed to construct a multi-model ensemble (MME) system. The ensemble reservoir inflow was simulated using the HBV hydrological model under the two modeling approaches involving deterministic forecasts (simulation using observed precipitation data as input) and ensemble forecasts (simulation using post-processed EPFs as input). The results demonstrated that both GMDH and WA-WLSR models had a positive impact on improving the forecast skill of the NWP models, but more accurate results were obtained by the WA-WLSR model. Ensemble forecasts outperformed coupled atmospheric–hydrological modeling in comparison with deterministic forecasts to simulate inflow hydrographs. Our proposed approach lends itself to quantifying uncertainty of ensemble forecasts in hydrometeorological the models, making it possible to have more reliable strategies for extreme-weather event management.

Exploiting multiple hydrologic forecasts to inform real-time reservoir operation for drought mitigation

To timely prompt anticipatory operations and relieve prolonged droughts effectively, multiple hydrological forecasts that provide thorough information about future streamflow need to be employed. However, it is challenging to integrate multiple forecasts in reservoir optimization and obtain a proper trading-off between current and future water supply benefits. This study proposes a novel Model Predictive Control (MPC) that can not only utilize the streamflow forecast but also utilize other two categorical forecasts, including the regime state forecast to capture the long-term persistence of the streamflow process and the annual streamflow volume state forecast. In the novel MPC, a discount factor depending on the annual streamflow volume state forecast weighs up the current deficit cost decided by the streamflow forecast and the future cost function determined by the regime state forecast. Taking the Biliuhe Reservoir in China as a case study, the value of each forecast in drought mitigation is evaluated. Results show that with an appropriate discount factor, using seasonal streamflow forecast can relieve intense deficits for extreme droughts and avoid unnecessary intense limitations for slight droughts. Utilizing the annual streamflow volume state forecast contributes to guaranteeing a reliable water supply. And employing the regime state forecast relieves the intensity of severe water deficit. By incorporating these hydrological forecasts, a 31.86% performance gain can be obtained with only a 5.03% reliability decrease compared to the baseline Stochastic Dynamic Programming informed by no forecast information. Nevertheless, forecast value is beset by forecast uncertainty. For streamflow forecast, forecast value decreases with increasing forecast uncertainty. But for the other two categorical forecasts, their forecast value depends on not only the forecast accuracy but also the hydrologic conditions and forecast bias.

A Hydrological Data Prediction Model Based on LSTM with Attention Mechanism

With the rapid development of IoT, big data and artificial intelligence, the research and application of data-driven hydrological models are increasing. However, when conducting time series analysis, many prediction models are often directly based on the following assumptions: hydrologic time series are normal, homogeneous, smooth and non-trending, which are not always all true. To address the related issues, a solution for short-term hydrological forecasting is proposed. Firstly, a feature test is conducted to verify whether the hydrological time series are normal, homogeneous, smooth and non-trending; secondly, a sequence-to-sequence (seq2seq)-based short-term water level prediction model (LSTM-seq2seq) is proposed to improve the accuracy of hydrological prediction. The model uses a long short-term memory neural network (LSTM) as an encoding layer to encode the historical flow sequence into a context vector, and another LSTM as a decoding layer to decode the context vector in order to predict the target runoff, by superimposing on the attention mechanism, aiming at improving the prediction accuracy. Using the experimental data regarding the water level of the Chu River, the model is compared to other models based on the analysis of normality, smoothness, homogeneity and trending of different water level data. The results show that the prediction accuracy of the proposed model is greater than that of the data set without these characteristics for the data set with normality, smoothness, homogeneity and trend. Flow data at Runcheng, Wuzhi, Baima Temple, Longmen Town, Dongwan, Lu’s and Tongguan are used as input data sets to train and evaluate the model. Metrics RMSE and NSE are used to evaluate the prediction accuracy and convergence speed of the model. The results show that the prediction accuracy of LSTM-seq2seq and LSTM-BP models is higher than other models. Furthermore, the convergence process of the LSTM-seq2seq model is the fastest among the compared models.

Application of robust deep learning models to predict mine water inflow: Implication for groundwater environment management

Traditional mine water inflow prediction is characterized by a high degree of uncertainty in model parameters and complex mechanisms involved in the water inflow process. Data-driven models play a key role in predicting inflow mechanisms without considering physical changes. However, the existing models are limited by nonlinearity and non-stationarity. Thus, the principal objective of this study was to propose two robust models, the DIFF-TCN model and the DIFF-LSTM model, for predicting the average water inflow per day. The models consist of three methods, namely Difference Method (DIFF), Temporal Convolutional Neural Network (TCN), and Long Short-Term Memory Neural Network (LSTM). When applied to the Tingnan Coal Mine, Shanxi Province, China, the DIFF-TCN performs better in predicting the average daily water inflow, the model has a MAE of 5.88 m3/h, RMSE of 6.85 m3/h and R2 of 0.96 in the test stage of the water inflow event. Comparison with the other deep learning models (with similar complex structures) and traditional time series model shows the superiority of our proposed DIFF-TCN model. The SHAP value is used to explain the contribution of each model input to the predicted values, and it indicates that the historical time of water inflow data are the most important input, and the advance distance and the groundwater level data also contribute to the model predictions, but groundwater level data for some periods in the past may have a detrimental effect on the model. The findings of this study can provide better understanding about potential of robust deep learning models for smart hydrological forecasting, and they can also provide technical guidance for mining safety production and protection of water resources and water environment around the mining area.

Pre- and postprocessing flood forecasts using Bayesian model averaging

Abstract In this study, pre- and postprocessing of hydrological ensemble forecasts are evaluated with a special focus on floods for 119 Norwegian catchments. Two years of ECMWF ensemble forecasts of temperature and precipitation with a lead time of up to 9 days were used to force the operational hydrological HBV model to establish streamflow forecasts. A Bayesian model averaging processing approach was applied to preprocess temperature and precipitation forecasts and for postprocessing streamflow forecasts. Ensemble streamflow forecasts were generated for eight schemes based on combinations of raw, preprocessed, and postprocessed forecasts. Two datasets were used to evaluate the forecasts: (i) all streamflow forecasts and (ii) forecasts for flood events with streamflow above mean annual flood. Evaluations based on all streamflow data showed that postprocessing improved the forecasts only up to a lead time of 2–3 days, whereas preprocessing temperature and precipitation improved the forecasts for 50–90% of the catchments beyond 3 days' lead time. We found large differences in the ability to issue warnings between spring and autumn floods. Spring floods had predictability for up to 9 days for many events and catchments, whereas the ability to predict autumn floods beyond 3 days was marginal.

Can Hydrological Models Benefit From Using Global Soil Moisture, Evapotranspiration, and Runoff Products as Calibration Targets?

AbstractHydrological models are usually calibrated to in‐situ streamflow observations with reasonably long and uninterrupted records. This is challenging for poorly gage or ungaged basins where such information is not available. Even for gaged basins, the single‐objective calibration to gaged streamflow cannot guarantee reliable forecasts because, as has been documented elsewhere, the inverse problem is mathematically ill‐posed. Therefore, the inclusion of other observations, and the reproduction of other hydrological variables beyond streamflow, become critical components of accurate hydrological forecasting. In this study, six single‐ and multi‐objective model calibration schemes based on different combinations of gaged streamflow, global‐scale gridded soil moisture, actual evapotranspiration (ET), and runoff products are used for the calibration of a process‐based hydrological model for 20 catchments located within the Lake Michigan watershed, of the Laurentian Great Lakes. Results show that the addition of gridded soil moisture to gaged streamflow in model calibration improves the ET simulation performance for most of the catchments, leading to the overall best‐performing models. The monthly streamflow simulation performance for the experiments using gridded runoff products to inform the model is outperformed by those using the gaged streamflow, but the discrepancy is mitigated with increasing catchment scale. A new visualization method that effectively synthesizes model performance for the simulations of streamflow, soil moisture, and ET was also proposed. Based on the method, it is revealed that the streamflow simulation performance is relatively weak for baseflow‐dominated catchments; overall, the 20 catchment models simulate streamflow and ET better than soil moisture.

Hydrological drought forecasting and monitoring system development using artificial neural network (ANN) in Ethiopia

The objective of this study is to investigate and perform long–term forecasting of both streamflow and hydrological drought over Ethiopia. Observed streamflow and precipitation data are collected from 17 streamflow stations and 34 rainfall gauge stations to forecast future streamflow and hydrological drought from 2026 to 2099. Streamflow forecasting is performed using an artificial neural network (ANN) in conjunction with python software. Observed precipitation and streamflow data from 1973 to 2014 are used to train and test the ANN model by 70 and 30% ratios, respectively. After training the model, future downscaled precipitation data from regional climate models (RCM) have been used as input data to forecast future streamflow. Three RCM models were used to downscale historical and future climate data. RACMO is found a good downscaling model for all selected stations. The linear scaling bias correction technique results in less than 2% error compared to other alternative techniques. The result indicates that ANN is a good tool to forecast streamflow in areas having a good correlation between precipitation and streamflow such as Abbay, Awash, Baro, Omo Gibe, and Tekeze river basins. But in arid areas for example Genale Dawa, Wabishebele, and Rift Valley basins, the model is not suitable because the input data (precipitation) have high variation than the output variable (streamflow). In such areas, meteorological drought analysis and forecasting are better than hydrological drought analysis. Finally, future hydrological drought is analyzed using forecasted streamflow data as input to the streamflow drought index (SDI). The result indicates that 2028, 2036, 2042, 2044, 2062, and 2063 are the expected extreme drought years in most river basins of Ethiopia in the future. This shows that at least one extreme drought is expected in each decade in the future. Therefore, extensive research in drought analysis and forecasting is needed to develop an effective drought early warning system, and water resource management policy.

The suitability of a seasonal ensemble hybrid framework including data-driven approaches for hydrological forecasting

Abstract. Hydrological forecasts are important for operational water management and near-future planning, even more so in light of the increased occurrences of extreme events such as floods and droughts. Having a forecasting framework, which is flexible in terms of input forcings and forecasting locations (local, regional, or national) that can deliver this information in fast and computational efficient manner, is critical. In this study, the suitability of a hybrid forecasting framework, combining data-driven approaches and seasonal (re)forecasting information from dynamical models, to predict hydrological variables was explored. Target variables include discharge and surface water levels for various stations at a national scale, with the Netherlands as the focus. Five different machine learning (ML) models, ranging from simple to more complex and trained on historical observations of discharge, precipitation, evaporation, and seawater levels, were run with seasonal (re)forecast data, including the European Flood Awareness System (EFAS) and ECMWF seasonal forecast system (SEAS5), of these driver variables in a hindcast setting. The results were evaluated using the evaluation metrics, i.e. anomaly correlation coefficient (ACC), continuous ranked probability (skill) score (CRPS and CRPSS), and Brier skill score (BSS), in comparison to a climatological reference hindcast. Aggregating the results of all stations and ML models revealed that the hindcasting framework outperformed the climatological reference forecasts by roughly 60 % for discharge predictions (80 % for surface water level predictions). Skilful prediction for the first lead month, independently of the initialization month, can be made for discharge. The skill extends up to 2–3 months for spring months due to snowmelt dynamic captured in the training phase of the model. Surface water level hindcasts showed similar skill and skilful lead times. While the different ML models showed differences in performance during a testing and training phase using historical observations, running the ML framework in a hindcast setting showed only minor differences between the models, which is attributed to the uncertainty in seasonal forecasts. However, despite being trained on historical observations, the hybrid framework used in this study shows similar skilful predictions to previous large-scale forecasting systems. With our study, we show that a hybrid framework is able to bring location-specific skilful seasonal forecast information with global seasonal forecast inputs. At the same time, our hybrid approach is flexible and fast, and as such, a hybrid framework could be adapted to make it even more interesting to water managers and their needs, for instance, as part of a fast model-predictive control framework.

Evaluation of parametric postprocessing of ensemble precipitation forecasts of the NCMRWF for the Vishwamitri River Basin

AbstractPostprocessing of the ensemble precipitation data improves the bias and uncertainty induced in the numerical weather prediction (NWP) due to perturbations of the initial condition of atmospheric models. The evaluation of the NWP of short-range quantitative weather forecasts provided by the NCMRWF archived in TIGGE, for the Vishwamitri River Basin. The aim of the study is to perform univariate statistical postprocessing using six parametric methods and compare to find the best suitable approach among censored non-homogeneous logistic regression (cNLR), Bayesian model averaging (BMA), logistic regression (logreg), heteroscedastic logistic regression (hlogreg), heteroscedastic extended logistic regression (HXLR), and ordered logistic regression (OLR) methods for the Vishwamitri River Basin. The Brier score (BS), the area under curve (AUC) of receiver operator characteristics (ROC) plots, Brier decomposition, and reliability plots were used for the verification of the probabilistic forecasts. In agreement with the BS and AUC, the cNLR approach for postprocessing performed very well for calibration at all five grids and is preferred, whereas BMA and hlogreg approaches showed relatively poor performance for the Vishwamitri basin. The best post-processed ensemble precipitation will further be employed as an input for the generation of the hydrological forecasts in operational flood forecasting in the Vishwamitri River Basin.

Connecting hydrological modelling and forecasting from global to local scales: Perspectives from an international joint virtual workshop

AbstractThe unprecedented progress in ensemble hydro‐meteorological modelling and forecasting on a range of temporal and spatial scales, raises a variety of new challenges which formed the theme of the Joint Virtual Workshop, ‘Connecting global to local hydrological modelling and forecasting: challenges and scientific advances’. Held from 29 June to 1 July 2021, this workshop was co‐organised by the European Centre for Medium‐Range Weather Forecasts (ECMWF), the Copernicus Emergency Management (CEMS) and Climate Change (C3S) Services, the Hydrological Ensemble Prediction EXperiment (HEPEX), and the Global Flood Partnership (GFP). This article aims to summarise the state‐of‐the‐art presented at the workshop and provide an early career perspective. Recent advances in hydrological modelling and forecasting, reflections on the use of forecasts for decision‐making across scales, and means to minimise new barriers to communication in the virtual format are also discussed. Thematic foci of the workshop included hydrological model development and skill assessment, uncertainty communication, forecasts for early action, co‐production of services and incorporation of local knowledge, Earth observation, and data assimilation. Connecting hydrological services to societal needs and local decision‐making through effective communication, capacity‐building and co‐production was identified as critical. Multidisciplinary collaborations emerged as crucial to effectively bring newly developed tools to practice.

Regional Adaptability of Global and Regional Hydrological Forecast System

Our paper aims to improve flood forecasting by establishing whether a global hydrological forecast system could be used as an alternative to a regional system, or whether it could provide additional information. This paper was based on the operational Global Flood Awareness System (GloFAS) of the European Commission Copernicus Emergency Management Service, as well as on a regional hydrological forecast system named RHFS, which was created with observations recorded in the Wangjiaba river basin in China. We compared the discharge simulations of the two systems, and tested the influence of input. Then the discharge ensemble forecasts were evaluated for lead times of 1–7 d, and the impact on the forecasts of errors in initialization and modelling were considered. We also used quantile mapping (QM) to post-process the discharge simulations and forecasts. The results showed: (1) GloFAS (KGE of 0.54) had a worse discharge simulation than RHFS (KGE of 0.88), mainly because of the poor quality of the input; (2) the average forecast skill of GloFAS (CRPSS about 0.2) was inferior to that of RHFS (CRPSS about 0.6), because of the errors in the initialization and the model, however, GloFAS had a higher forecast quality than RHFS at high flow with longer lead times; (3) QM performed well at eliminating errors in input, the model, and the initialization.

A Decomposition-based Multi-model and Multi-parameter ensemble forecast framework for monthly streamflow forecasting

Hydrological ensemble forecasting plays a critical role in decision-making and water resource management. The skill of an ensemble forecasting system is limited by input data, model parameters and model structure, and the precision of the system depends mainly on the ensemble size. The main contradiction is to ensure the forecast performance and ensemble size under considering the uncertainties of inputs and models. In this study, a hybrid decomposition-based multi-model and multi-parameter (DMP) ensemble streamflow forecast method is proposed. The proposed method couples the signal decomposition and Artificial Intelligence (AI) forecast method in order to provide a more efficient and effective streamflow forecast. The decomposition method is used to reduce the uncertainty of inputs for the forecast model, and AI models are used to improve the performance of forecasting. The results illustrate that the DMP ensemble streamflow forecast method not only extracts the characteristic periodic term and trend term of hydrological series but also improves the forecasting accuracy and reduces the ensemble forecast uncertainty. At the same time, it greatly expands the ensemble size which solves the problem of insufficient ensemble size in order to be convenient for further water resources analysis and decision-making. The findings also show that high-frequency subseries are highly sensitive to ensemble forecast, and it is recommended that more than 2 layers of high-frequency series should be used for ensemble forecast. The proposed approach in this study is more suitable for nonlinear and non-stationary hydrological series forecasting, and it is of reference significance for real reservoir operation.

Multi-step ahead probabilistic forecasting of multiple hydrological variables for multiple stations

The demand for more accurate simulation of physical river basin puts forward higher requirements on the number of hydrological forecast stations, types of hydrological variables, forecast accuracy, forecast period and quantitative uncertainty. Therefore, how to simultaneously obtain multi-step ahead probabilistic forecasting of multiple hydrological variables for multiple stations is a key issue to be solved in this study. Firstly, the background of River Basin Digital Twin is introduced and the new problems faced by hydrological forecasting in this background are analyzed. Then, the input and output of hydrological forecasting are reconstructed into 4-D tensors. Next, a new hybrid deep learning model (B-CM-C3D) based on 3-D Convolutional Neural Network, Convolutional Minimum Gate Memory Neural Network and Variational Bayesian Neural Network is proposed to obtain multi-step ahead probabilistic forecasting of multiple hydrological variables for multiple stations. Finally, in order to verify the performance of this model, it was compared with four state-of-the-art models in the Yangtze River Basin. The experimental results show that: (1) the deterministic prediction accuracy and probabilistic forecasting comprehensive performance of B-CM-C3D are better than other comparison models in 80% of the results. (2) B-CM-C3D shortens the training time of B-CL-C3D by 43% and improves the prediction accuracy.

Possibilities of application of logistic regression in hydrological forecasts (on the example of the mountain river Samur)

The possibilities of constructing regression equations for predicting the runoff of mountain and semi-mountain rivers are considered. A predictive equation for the Samur river catchment has been obtained, which connects water discharge and predictors by casual relationships: water levels, air temperature, air humidity, atmospheric pressure, precipitation, dew point temperature, wind direction, and cloudiness. Logistic regressions are obtained, which allow using categorical variables as independent variables. The result of a logistic regression forecast is the probability of the occurrence or non-occurrence of the event of the predicted value. The positive and negative aspects of this approach for mountain rivers are revealed, which consist of the interpretation of the predicted probability of the event. Actions are proposed that allow for obtaining more reliable forecasts.

ДОВГОСТРОКОВЕ ПРОГНОЗУВАННЯ ДАТИ ПОЧАТКУ ВЕСНЯНОГО ВОДОПІЛЛЯ У ВЕРХІВ’Ї РІЧКИ ПІВДЕННИЙ БУГ ЗА ТЕЛЕКОННЕКЦІЙНИМИ ІНДЕКСАМИ

The start date of spring flood is an important hydrological characteristic. Insufficient attention is paid to its long-term forecasting, which is due to the complexity and unsolved problem in terms of improving the quality of such forecasting. Most often, quantitative methods are used in long-term forecasting. The most used are statistical, correlation, and regression analysis. Recently, teleconnection indices and patterns are increasingly used in hydrological long-term forecasting. At the same time, the basis of the concept of forecasting by the teleconnection connections is the idea of the influence of distant fluctuations of atmospheric circulation on the hydrological event. So, the theteleconnection indices and patterns are used for forecasting and analysis of river flow, atmospheric precipitation, research of snow water equivalent of river basins, forecasting of droughts and ice phenomena. The objective of this of the study is to develop a methodology of the long-term forecast of the spring flood start date in the upper part of the Southern Buh River using teleconnection indices and patterns. The method of long-term forecasting of the start date of spring flood was developed for the Southern Buh River – Lelitka village water gauge. The Southern Buh River – Lelitka village water gauge is located in the forest zone and characterizes homogeneous conditions of the formation of spring flood. Information on the start dates of spring flood for the observation period 1966-2015 and teleconnection indices and patterns were used. The start dates of spring flood at the Southern Buh River near Lelitka village are characterized by significant variability. So, the difference between late and early dates of spring flood is 65 days. 34 teleconnection indices and patterns were used, which the National Oceanic & Atmospheric Administration USA were determined. The best regression relationship with the start dates of spring flood at the South Buh River – Lelitka village water gauge was obtained for the indices WPAC850 in January and AAO in December. The technique corresponds to the “satisfactory” category for the probability of not exceeding the permissible error, which allows it to be used for forecasting. So, the teleconnection indices and patterns can be quite successfully used in the long-term forecasting of the start date of spring flood.

Artificial Intelligence Technique in Hydrological Forecasts Supporting for Water Resources Management of a Large River Basin in Vietnam

Artificial Intelligence Technique in Hydrological Forecasts Supporting for Water Resources Management of a Large River Basin in Vietnam