Dam Inflow Research Articles

Unraveling the Relationships Between Trend of Dam Inflows, Hydrometeorological Variables and Vegetation in West and Southwest United States

Abstract This paper explored temporal changes in magnitude and seasonality of low, median, and high inflows of 51 dams across the West and Southwest United States over the 1993-2022 period. Changes in precipitation, air temperature (an indicator of snowpack and evaporation), soil moisture, and vegetation were also examined to identify potential reasons for the temporal trends in dam inflows. Using monotonic and non-monotonic tests, we found a general downward trend in dam inflows, particularly across the Upper Colorado and California regions. More than 30% of the dams showed a downward trend in their annual median inflows, high inflows during spring, and median inflows during fall. The downward trend of dam inflows was associated with decreasing precipitation and soil moisture, and rising temperatures. While vegetation exhibited positive associations with inflows, it did not seem to be a primary factor for explaining the inflow trends. We also observed shifts in the seasonality of low and high inflows; there was an increase in the proportion of inflows occurring during summer and fall, and a decrease in winter proportions for low inflows. Similarly, high inflows exhibited an increase in spring proportions and a decrease in fall proportions. Our changepoint analyses detected non-monotonic trends between 2002 and 2012 in ~13% of the dams; the majority were located in the Upper Colorado and California regions. More than half of these changepoints were in 2011, likely due to widespread droughts then. Our study has implications for reservoir managers to identify changes that dams experience over time and assist them in proposing actions that maintain the dams’ functionality.

Evaluating machine learning models in predicting dam inflow and hydroelectric power production in multi-purpose dams (case study: Mahabad Dam, Iran)

This study aimed to forecast dam inflows and subsequently predict its capability in producing HEPP using machine learning and evolutionary optimization techniques. Mahabad Dam, located in the northwest of Iran and recognized as one of the nation’s key dams, served as a case study. First, artificial neural networks (ANN) and support vector regression (SVR) were employed to predict dam inflows, with optimization of parameters achieved through Harris hawks optimization (HHO), a robust optimization technique. The data of temperature, precipitation, and dam inflow over a 24-year period on a monthly basis, incorporating various lag times, were used to train these machines. Then, HEPP from the dam was predicted using temperature, precipitation, dam inflow, and dam evaporation as input variables. The models were applied to data covering the years 2000 to 2020. The results of the first part indicated both hybrid models (HHO-ANFIS and HHO-SVR) improved the prediction performance compared to the single models. Based on the results of Taylor’s diagram and the error evaluation criteria, the HHO-ANFIS hybrid model (RMSE, MAE, and NSE of 3.90, 2.41, and 0.86, respectively) exerted better performance than HHO-SVR (RMSE, MAE, and NSE of 4.39, 2.70, and 0.86, respectively). The results of the second part showed that using the HHO algorithm to optimize single models (RMSE, MAE, and NSE of 0.2, 10, and 0.90, respectively) predicted HEPP better than single models (RMSE, MAE, and NSE of 0.2, 10, and 0.90, respectively). The results of Taylor’s diagram also showed that the HHO-ANFIS model exerted better performance. The findings of this study indicated the promising performance of machine learning models optimized by metaheuristic algorithms in the simultaneous prediction of dam inflows and HEPP in multi-purpose dams for better management and allocation of surface water resources.

Uncertainty evaluation of dam inflow under the influence of climate change (study area: Latiyan Dam)

Uncertainty evaluation of dam inflow under the influence of climate change (study area: Latiyan Dam)

Predicting Hydrological Drought Conditions of Boryeong Dam Inflow Using Climate Variability in South Korea

Predicting Hydrological Drought Conditions of Boryeong Dam Inflow Using Climate Variability in South Korea

Assessing the effectiveness of artificial intelligence models in predicting Zayanderud dam inflow at different time scales

Assessing the effectiveness of artificial intelligence models in predicting Zayanderud dam inflow at different time scales

Trend detection and change point analysis of inflows in Karuppanadhi and Gundar dams of Chittar River Basin, Tamil Nadu, India

Abstract Gaining insights into the patterns of hydroclimatic elements such as inflows to dams is essential to adeptly strategize and oversee water resource planning. For this study, we focused on Karuppanadhi and Gundar dams within the Chittar River Basin in Tamil Nadu. We employed various time series trend assessment methods – Mann–Kendall (MK), modified Mann–Kendall (MMK), Sen slope estimator, and innovative trend analysis (ITA). Furthermore, we conducted change point detection analysis using homogeneity tests, namely Pettitt's test, standard normal homogeneity test (SNHT), and Buishand test. The analysis was carried out across three-time scales: monthly, seasonal, and annual for 30 years (1991–2020) span. Results revealed declining trends significantly across all three timescales for the Karuppanadhi dam. Whereas for the Gundar dam, notable trends included increased patterns in January, March, December, and the winter season, while other months/seasons showed statistically decreasing trends. ITA exhibited greater sensitivity in identifying trends at monthly and seasonal scales, indicating its superior trend detection over MK and MMK methods. Change point analysis of trends identified a rising trend in August post-2017 for both dams. Other change points indicated decreasing inflow trends thereafter. Analysing trends and change points in dam inflows aid decisions on water resources management.

Water quantity–quality assessment in the transboundary river basin under climate change: a case study

Abstract This study combined hydrological and water quality simulation models with a water resources planning model to project future water supply conditions under the dam construction in the Harirud River, located at the Afghanistan, Turkmenistan, and the Iran border. The sustainability requirements and possible conflicts among riparian countries were assessed under climate change and future development in Afghanistan's upstream. The water quantity and quality of the Doosti Dam Basin on the Harirud River were investigated based on a contemporary time (1955–2015) to predict the future condition (2020–2099). The representative concentration pathway scenarios were applied based on five bias-corrected climate models. Results showed that most areas of the study area experienced an increase in temperature (1.5–3.8°C) and a decrease in precipitation (19–24%). The Doosti Dam inflow decreased by about 70% after the Salma Dam construction, and the reliability and sustainability of agricultural water supply in Iran and Turkmenistan will reduce to less than 3% under the representative concentration pathways 8.5 climate change scenario. The results show that the Doosti Dam is not a reliable source to supply the domestic water demand of Mashhad, the second most important city in Iran.

Climate stress impacts on the reservoir inflows: a decision-scaling and IHACRES modeling approach in South Korean basins

Abstract This study employed the Decision-Scaling (DS) approach and the IHACRES (Identification of unit Hydrographs And Component flows from Rainfall) long-term runoff model to investigate the impact of composite changes in rainfall and temperature on dam inflows at Chungju dam, Yongdam dam, Hapcheon dam, and Seomjingang dam basins. By analyzing flow regimes, rainfall scenarios, and temperature scenarios, the study revealed crucial insights into dam inflow and its response to climate stressors. The findings demonstrated that the rate of inflow increase in the study basins exceeded the rate of rainfall increase, indicating the significance of basin storage effects in contributing to runoff generation. The analysis of rainfall changes by quantile highlighted the dominant influence of the upper-high (UH) rainfall quantile, which led to higher flood inflow compared to other quantiles. Furthermore, scenarios with different rainfall patterns were compared, showcasing the predominant impact of UH quartile rainfall on the total inflow variation. The study also analyzed the runoff ratio, finding that changes in precipitation proportionally affected the runoff ratio. Overall, these insights contribute to understanding the sensitivity of dams to changes in rainfall and temperature, facilitating the development of strategies for sustainable water supply and flood management in dam systems.

Comparative Study of Machine Learning and Deep Learning Models Applied to Data Preprocessing Methods for Dam Inflow Prediction

In this study, we employed representative machine learning (ML) and deep learning (DL) models previously utilized in the fields of rainfall and runoff analysis in the water resources sector. We not only performed hyperparameter tuning of the models but also considered the characteristics of the model and the combination and preprocessing (such as lag-time and moving average) of meteorological and hydrological data. We then compared and evaluated the performance of the models according to various scenarios of data characteristics and ML & DL model combinations for predicting daily water inflow. To accomplish this, we utilized meteorological and hydrological data collected from 1974 to 2021 in the Soyang River Dam Basin to examine 1) precipitation, 2) inflow, and 3) meteorological data as primary independent variables. We then employed a total of 36 scenario combinations as input data for ML & DL, applying a) lag-time, b) moving average, and c) component separation conditions for inflow. To identify the most suitable data combination characteristics and ML & DL models for predicting daily inflow, we compared and evaluated 10 different ML & DL models: 1) Linear Regression, 2) Lasso, 3) Ridge, 4) Support Vector Regression, 5) Random Forest (RF), 6) Light Gradient Boosting Model, 7) XGBoost for ML, and 8) Long Short-Term Memory (LSTM) models, 9) Temporal Convolutional Network (TCN), and 10) LSTM-TCN for DL.

Development of Multi-Inflow Prediction Ensemble Model Based on Auto-Sklearn Using Combined Approach: Case Study of Soyang River Dam

Accurate prediction of dam inflows is essential for effective water resource management and dam operation. In this study, we developed a multi-inflow prediction ensemble (MPE) model for dam inflow prediction using auto-sklearn (AS). The MPE model is designed to combine ensemble models for high and low inflow prediction and improve dam inflow prediction accuracy. We investigated the impact of datasets assigned to flow regimes on the ensemble composition and compared the performance of the MPE model to an AS-based ensemble model developed using a conventional approach. Our findings showed that the MPE model outperformed the conventional model in predicting dam inflows during flood and nonflood periods, reducing the root mean square error (RMSE) and mean absolute error (MAE) by 22.1% and 24.9% for low inflows, and increasing the coefficient of determination (R2) and Nash–Sutcliffe efficiency (NSE) by 21.9% and 35.8%, respectively. These results suggest that the MPE model has the potential to improve water resource management and dam operation, benefiting both the environment and society. Overall, the methodology of this study is expected to contribute to the development of a robust ensemble model for dam inflow prediction in regions with high climate variability.

Generation of Synthetic Series for Long-Term Analysis of Optimal Operation Policies of a Cascade Hydroelectric Dam System

Stochastic Dynamic Programming (SDP) has been used to solve reservoir management problems in different parts of the world; specifically in Mexico, it has been used to obtain operating policies that optimize a given objective function. By simulating the operation of the system with a comprehensive model, the behavior of such policies can be accurately evaluated. An optimal policy involves, on the one hand, the selection of the volume of water to extract from each reservoir of the system that guarantees the maximum expected benefit from electricity generation in the long term; and, on the other hand, an optimal policy should reduce the occurrence of unwanted events such as spills, deficits, as well as volumes exceeding the guide curves imposed by the operators of the dams. In the case of the Grijalva river dam system, SDP was applied to determine optimal operating policies considering three alternative guide curves proposed by different agencies; however, since the simulation of the operation of the system under the three alternatives with the historical record of dam inflows found that none of them showed deficits or spills, it was considered necessary to use synthetic series of inflows to increase the stress of the system. Records of synthetic biweekly series of 1000 years were then generated to simulate the behavior of the Grijalva river dam system using the optimal operation policies obtained for each alternative. By stressing the dam system by simulating its behavior with synthetic series longer than the historical record but preserving the same statistical characteristics of the historical series on the synthetic ones, it was possible to realistically evaluate each operating policy considering the frequency and magnitude of spills and deficits that occurred at each dam. For the generation of the synthetic series, a fragment method was used; it was adapted to simultaneously generate the inflow volumes to the two regulating dams (modified Svanidze method), which preserves the statistical characteristics of the historical series, including both the autocorrelations of each series and the cross-correlation. It was also verified that simulating the operation of the dam system with the generated series also preserves the average conditions, such as the average biweekly generation at each dam, which were obtained in the simulations with the historical record. Finally, an optimal policy was obtained (Test 4) by combining the guide curves used in the previous tests. Such a policy attained an average energy production of 474 GWh/fortnight, the lowest average total spills in the system (30,261.93 hm3), and limited deficits (5973.17 hm3) in the long term. This represents a relative increase of 16% in energy generated compared to the balanced historical operation scenario with respect to the few events of spills and deficits.

Dam inflow prediction using large-scale climate variability and deep learning approach: a case study in South Korea

AbstractAccurate prediction of dam inflows is essential for effective water resources management in terms of both water quantity and quality. This study aims to develop a Long Short-Term Memory (LSTM) deep learning-based monthly dam inflow prediction model using large-scale climate indices. Six climate indices, Atlantic multidecadal oscillation (AMO), El Niño–southern oscillations (ENSO), North Atlantic oscillation (NAO), Pacific decadal oscillation (PDO), Niño 3.4, and Southern Oscillation Index (SOI) for the period of 1981–2020, were used as input variables of the model. The proposed model was trained with 29 years of data (1981–2009) and tested with 12 years of data (2009–2020). We investigated 29 input data combinations to evaluate the predictive performance according to different input datasets. The model showed the average values of metrics ranged from 0.5 to 0.6 for CC and from 40 to 80 cm for root mean square error (RMSE) at three dams. The prediction results from the model showed lower performance as the lead time increased. Also, each dam showed different prediction results for different seasons. For example, Soyangriver/Daecheong dams have better accuracy in prediction for the wet season than the dry season, whereas the Andong dam has a high prediction ability during the dry season. These investigations can be used for better efficient dam management using a data-driven approach.

Rainfall Spatial-Temporal Variability and Trends in the Thamirabharani River Basin, India: Implications for Agricultural Planning and Water Management

Rainfall is critical to agricultural and drinking water supply in the Thamirabharani river basin. The upper catchment areas of the Thamirabharani basin are located in high-elevated forest regions, and rainfall variability affects dam inflow and outflow. The well-known methods for rainfall analysis such as the coefficient of variation (CV), the precipitation concentration index (PCI), and trend analysis by Mann-Kendall and Sen’s slope test, as well as the Sen’s graphical innovative trend method (ITA) recently reported in several studies, were used. Rainfall data from gauge stations and the satellite-gridded Multisource Weighted Ensemble Precipitation (MSWEP) dataset were chosen for analysis at the annual and four-season time scales, namely, the Southwest Monsoon, Northeast Monsoon, winter, and summer seasons from 1991 to 2020. The mean annual PCI value reflects irregular monthly rainfall distribution (PCI > 20) in all gauge stations. The spatial monthly rainfall distribution of PCI values remarkedly shows a moderate distribution in the western and an anomalous distribution in the eastern part of the basin. The annual mean rainfall ranges from 718.4 to 2268.6 mm/year, decreasing from the high altitude zone in the west to the low plains and coastal regions in the east. Seasonal rainfall contributes about 42% from the NEM, 30.6% from the SWM, 22.8% from summer, and 3.9% from winter, with moderate variability (CV less than 30%). Ground stations experienced extremely high interannual variability in rainfall (more than 60%). Trend analysis by the MK, TFPW-MK, and ITA methods shows increasing annual rainfall in the plains and coastal regions of the basin; particularly, more variations among the seasons were observed in the Lower Thamirabharani sub-basin. The NEM and summer season rainfall are statistically significant and contribute to the increasing trend in annual rainfall. The ITA method performed better in the annual and seasonal scale for detecting the rainfall trend than the MK and TFPW-MK test. The Lower Thamirabharani sub-basin in the eastern part of the basin receives more rain during the NEM than in other areas. To summarize, the low plains in the central and coastal regions in the southeast part experience an increase in rainfall with irregular monthly distribution. This study helps farmers, governments, and policymakers in effective agricultural crop planning and water management.

A Study on the Optimal Deep Learning Model for Dam Inflow Prediction

In the midst of climate change, the need for accurate predictions of dam inflow to reduce flood damage along with stable water supply from water resources is increasing. In this study, the process and method of selecting the optimal deep learning model using hydrologic data over the past 20 years to predict dam inflow were shown. The study area is Andong Dam and Imha Dam located upstream of the Nakdong River in South Korea. In order to select the optimal model for predicting the inflow of two dams, sixteen scenarios (2 × 2 × 4) are generated considering two dams, two climatic conditions, and four deep learning models. During the drought period, the RNN for Andong Dam and the LSTM for Imha Dam were selected as the optimal models for each dam, and the difference between observations was the smallest at 4% and 2%, respectively. In typhoon conditions, the GRU for Andong Dam and the RNN for Imha Dam were selected as optimal models. In the case of Typhoon Maemi, the GRU and the RNN showed a difference of 2% and 6% from the observed maximum inflow, respectively. The optimal recurrent neural network-based models selected in this study showed a closer prediction to the observed inflow than the SFM, which is currently used to predict the inflow of both dams. For the two dams, different optimal models were selected according to watershed characteristics and rainfall under drought and typhoon conditions. In addition, most of the deep learning models were more accurate than the SFM under various typhoon conditions, but the SFM showed better results under certain conditions. Therefore, for efficient dam operation and management, it is necessary to make a rational decision by comparing the inflow predictions of the SFM and deep learning models.

Conjunctive Operation of Sand Dam and Groundwater Well for Reliable Water Supply during Drought Conditions

Some mountainous regions without water service facilities are among the areas most vulnerable to drought. In these locations, it is particularly essential to establish practical alternatives to cope with the increase in the intensity and duration of droughts caused by climate change. This study proposes a methodology for the conjunctive use of a sand dam and groundwater well under various drought conditions. The method has been applied to a small mountainous area in South Korea. Owing to the scarcity of observational data, it is crucial to properly estimate the hydrological components necessary for judging the feasibility and reliability of conjunctive operations. The step-by-step procedures for performing the tasks are presented in this study. For the inflow of the sand dam, which is a portion of the basin runoff, two different approaches were employed and compared: the Kajiyama formula and a simple two-parameter monthly water balance model (TPM). Water budget analysis allowed for the determination of whether the current and increased water demand could be met under various drought conditions. Preliminary analysis revealed that a sand dam alone could not reliably meet the demand for 10-year or more severe drought conditions. Various water allocation scenarios between surface water (i.e., sand dam) and groundwater were tested. Conjunctive use of a sand dam and groundwater well turned out to increase the reliability of the water supply. As water demand increases and droughts become more severe, the role of groundwater increases. With appropriate resource allocation, 100% water supply reliability could be achieved, even for one year-lasting 50-year drought. We demonstrated how a system could be flexibly operated to meet the target demands monthly, given the system reliability level.

Development of a Revised Multi-Layer Perceptron Model for Dam Inflow Prediction

It is necessary to predict dam inflow in advance for flood prevention and stable dam operations. Although predictive models using deep learning are increasingly studied, these existing studies have merely applied the models or adapted the model structure. In this study, data preprocessing and machine learning algorithms were improved to increase the accuracy of the predictive model. Data preprocessing was divided into two types: The learning method, which distinguishes between peak and off seasons, and the data normalization method. To search for a global solution, the model algorithm was improved by adding a random search algorithm to the gradient descent of the Multi-Layer Perceptron (MLP) method. This revised model was applied to the Soyang Dam Basin in South Korea, and deep learning-based discharge prediction was performed using historical data from 2004 to 2021. Data preprocessing improved the accuracy by up to 61.5%, and the revised model improved the accuracy by up to 40.3%. With the improved algorithm, the accuracy of dam inflow predictions increased to 89.4%. Based on these results, stable dam operation is possible through more accurate inflow predictions.

Assessment of Termination Criteria at Each Drought Response Stage on Climate Change in a Multi-Purpose Dam

In this study, the termination criteria at each drought response stage were shown to increase storage volume while maintaining the stability of the water supply service during drought in multi-purpose dams. For 52 of the Representative Concentration Pathway (RCP) 4.5/8.5 scenarios of the IPCC (Intergovernmental Panel on Climate Change) fifth assessment report (AR5), dam inflows were calculated through the precipitation-runoff modeling system (PRMS), and time reliability and supply reliability were estimated. Moreover, given the application of termination criteria at each drought response stage, the volume of additional water supply was calculated, while the number of days for additional supply availability of water for residential and industrial use and water for residential, industrial, and agricultural use and river maintenance was quantified as well. The CMCC-CMS(RCP4.5) GCM exhibited the largest volume of additional water supply at 74.15 million m3, which corresponds to 52.0 days of water for residential and industrial use and to 47.7 days of water for residential, industrial, and agricultural use and river maintenance. The analysis revealed that the volume and the number of days of additional water supply also increased for other GCMs due to the application of termination criteria at each drought response stage.

Unfolding the relationship between seasonal forecast skill and value in hydropower production: a global analysis

Abstract. The potential benefits of seasonal streamflow forecasts for the hydropower sector have been evaluated for several basins across the world but with contrasting conclusions on the expected benefits. This raises the prospect of a complex relationship between reservoir characteristics, forecast skill, and value. Here, we unfold the nature of this relationship by studying time series of simulated power production for 735 headwater dams worldwide. The time series are generated by running a detailed dam model over the period 1958–2000 with three operating schemes: basic control rules, perfect forecast-informed operations, and realistic forecast-informed operations. The realistic forecasts are issued by tailored statistical prediction models – based on lagged global and local hydroclimatic variables – predicting seasonal monthly dam inflows. As expected, results show that most dams (94 %) could benefit from perfect forecasts. Yet, the benefits for each dam vary greatly and are primarily controlled by the time-to-fill value and the ratio between reservoir depth and hydraulic head. When realistic forecasts are adopted, 25 % of dams demonstrate improvements with respect to basic control rules. In this case, the likelihood of observing improvements is controlled not only by design specifications but also by forecast skill. We conclude our analysis by identifying two groups of dams of particular interest: dams that fall in regions expressing strong forecast accuracy and having the potential to reap benefits from forecast-informed operations and dams with a strong potential to benefit from forecast-informed operations but falling in regions lacking forecast accuracy. Overall, these results represent a first qualitative step toward informing site-specific hydropower studies.

Monitoring Inflow Dynamics in a Multipurpose Dam Based on Travel-time Principle

Understanding inflow dynamics in a dam lake forms the basis for optimal dam operation and management practices. However, methods pertaining to adequately determining negative inflows and addressing them, as well as quantifying uncertainties in dam inflow, have been scarcely investigated. In this study, the inflow was observed using two pairs of fluvial acoustic tomography (FAT) systems placed diagonally in a dam lake, forming a crossed-shaped pattern. The “travel-time” principle is the primary approach for measuring the inflow by FAT. The novelty of this study is in discussing the inflow characteristics within a slow water-flow environment monitored by FAT. Based on the reciprocal sound transmission, we upgraded an equation to estimate the flow direction; this newly proposed generalized equation can be used in a fluctuating flow environment. We also discussed the sound propagation characteristics for slow flow velocities. Finally, we demonstrated that a small inaccuracy in the acoustic signal, even by a sub-millisecond, can cause significant errors in measurements. One of the novel findings of this study is the detection of internal waves using the improved flow direction equation and acoustic travel-time records. Overall, this study presents a promising approach for inflow measurements under extremely slow flow conditions.Supplementary InformationThe online version contains supplementary material available at 10.1007/s11269-022-03161-w.

Response of Fish Community to Building Block Methodology Mimicking Natural Flow Regime Patterns in Nakdong River in South Korea

Water regulation and flood control of rivers are changing due to streamflow depletion following industrialization and urbanization, significantly impacting aquatic ecosystems. Therefore, restoration of the ecological environment is necessary to maintain a healthy river ecosystem. For ecosystem restoration, the amount of discharge from dams must be controlled and the appropriate environmental flow must be calculated according to the fish species. The change in the flow through the dam due to hydropeaking directly impacts the fish. This study aimed to construct a building block methodology (BBM) using dam inflows in the Gudam Bridge basin upstream of the Nakdong River, build a River2D model of this area, and calculate the natural flow regime and the weighted usable area (WUA). The analysis of the scenarios for the whole period (2006–2020) and by flow regime showed that WUA decreased in some periods, but improved overall in the scenario reflecting the BBM. For Zacco platypus, a dominant fish species of the Gudam Bridge, WUA decreased by ~11% in some periods (in September) but the habitat improvement effect measured up to 79%. Changing the dam discharge pattern by considering the flow regime seemed more effective in improving the habitat of fish living downstream.