Deep Learning Prediction of Streamflow in Portugal
Hydrology
Abstract and Figures
The transformative potential of deep learning models is felt in many research fields, including hydrology and water resources. This study investigates the effectiveness of the Temporal Fusion Transformer (TFT), a deep neural network architecture for predicting daily streamflow in Portugal, and benchmarks it against the popular Hydrologiska Byråns Vattenbalansavdelning (HBV) hydrological model. Additionally, it evaluates the performance of TFTs through selected forecasting examples. Information is provided about key input variables, including precipitation, temperature, and geomorphological characteristics. The study involved extensive hyperparameter tuning, with over 600 simulations conducted to fine–tune performances and ensure reliable predictions across diverse hydrological conditions. The results showed that TFTs outperformed the HBV model, successfully predicting streamflow in several catchments of distinct characteristics throughout the country. TFTs not only provide trustworthy predictions with associated probabilities of occurrence but also offer considerable advantages over classical forecasting frameworks, i.e., the ability to model complex temporal dependencies and interactions across different inputs or weight features based on their relevance to the target variable. Multiple practical applications can rely on streamflow predictions made with TFT models, such as flood risk management, water resources allocation, and support climate change adaptation measures.
Figures - available from: Hydrology
This content is subject to copyright.
This study explores the application of Temporal Fusion Transformers (TFTs) to improve the predictability of hourly potential hydropower production for a small run–of–the–river hydropower plant in Portugal. Accurate hourly power forecasts are essential for optimizing participation in the spot electricity market, where deviations incur penalties. This research introduces the novel application of the TFT, a deep–learning model tailored for time series forecasting and uncovering complex patterns, to predict hydropower production based on meteorological data, historical production records, and plant capacity. Key challenges such as filtering observed hydropower outputs (to remove strong, and unpredictable human influence) and adapting the historical series to installed capacity increases are discussed. An analysis of meteorological information from several sources, including ground information, reanalysis, and forecasting models, was also undertaken. Regarding the latter, precipitation forecasts from the European Centre for Medium–Range Weather Forecasts (ECMWF) proved to be more accurate than those of the Global Forecast System (GFS). When combined with ECMWF data, the TFT model achieved significantly higher accuracy in potential hydropower production predictions. This work provides a framework for integrating advanced machine learning models into operational hydropower scheduling, aiming to reduce classical modeling efforts while maximizing energy production efficiency, reliability, and market performance.
Drought monitoring and prediction have important roles in various aspects of hydrological studies. In the current research, the standardized precipitation index (SPI) was monitored and predicted in Peru between 1990 and 2015. The current study proposed a hybrid model, called ANN-FA, for SPI prediction in various time scales (SPI3, SPI6, SPI18, and SPI24). A state-of-the-art firefly algorithm (FA) has been documented as a powerful tool to support hydrological modeling issues. The ANN-FA uses an artificial neural network (ANN) which is coupled with FA for Lima SPI prediction via other stations. Through the intelligent utilization of SPI series from neighbors’ stations as model inputs, the suggested approach might be used to forecast SPI at various time scales in a meteorological station with insufficient data. To conduct this, the SPI3, SPI6, SPI18, and SPI24 were modeled in Lima meteorological station using other meteorological stations’ datasets in Peru. Various error criteria were employed to investigate the performance of the ANN-FA model. Results showed that the ANN-FA is an effective and promising approach for drought prediction and also a multi-station strategy is an effective strategy for SPI prediction in the meteorological station with a lack of data. The results of the current study showed that the ANN-FA approach can help to predict drought with the mean absolute error = 0.22, root mean square error = 0.29, the Pearson correlation coefficient = 0.94, and index of agreement = 0.97 at the testing phase of best estimation (SPI3).
The application of machine learning (ML) and remote sensing (RS) in soil and water conservation has become a powerful tool. As analytical tools continue to advance, the variety of ML algorithms and RS sources has expanded, providing opportunities for more sophisticated analyses. At the same time, researchers are required to select appropriate technologies based on the research objectives, topic, and scope of the study area. In this paper, we present a comprehensive review of the application of ML algorithms and RS that has been implemented to advance research in soil and water conservation. The key contribution of this review paper is that it provides an overview of current research areas within soil and water conservation and their effectiveness in improving prediction accuracy and resource management in categorized subfields, including soil properties, hydrology and water resources, and wildfire management. We also highlight challenges and future directions based on limitations of ML and RS applications in soil and water conservation. This review aims to serve as a reference for researchers and decision-makers by offering insights into the effectiveness of ML and RS applications in the fields of soil and water conservation.
The application of machine learning (ML) and remote sensing (RS) in soil and water conserva-tion has become a powerful tool. As analytical tools continue to advance, the variety of ML al-gorithms and RS sources has expanded, providing opportunities for more sophisticated analyses. At the same time, researchers are required to select appropriate technologies based on research objectives, topic, and scope of the study area. In this paper, we present a comprehensive review on the application of ML algorithms and RS that has been implemented to advance research in soil and water conservation. The key contribution of this review paper is that it provides an overview of current research areas within soil and water conservation and their effectiveness in improving prediction accuracy and resource management in categorized subfields, including biomass-vegetation, soil properties, hydrology and water resources, and wildfire management. We also highlight challenges and future directions based on limitations of ML and RS applica-tions in soil and water conservation. This review aims to serve as a reference for researchers and decision-makers by offering insights into the effectiveness of ML and RS applications in the field of soil and water conservation.
The forecasting of hydrological flows (rainfall depth or rainfall discharge) is becoming increasingly important in the management of hydrological risks such as floods. In this study, the Long Short-Term Memory (LSTM) network, a state-of-the-art algorithm dedicated to time series, is applied to predict the daily water level of Lake Nokoué in Benin. This paper aims to provide an effective and reliable method to enable the reproduction of the future daily water level of Lake Nokoué, which is influenced by a combination of two phenomena: rainfall and river flow (runoff from the Ouémé River, the Sô River, the Porto-Novo lagoon, and the Atlantic Ocean). Performance analysis based on the forecasting horizon indicates that LSTM can predict the water level of Lake Nokoué up to a forecast horizon of t + 10 days. Performance metrics such as Root Mean Square Error (RMSE), coefficient of correlation (R²), Nash–Sutcliffe Efficiency (NSE), and Mean Absolute Error (MAE) agree on a forecast horizon of up to t + 3 days. The values of these metrics remain stable for forecast horizons of t + 1 day, t + 2 days, and t + 3 days. The values of R² and NSE are greater than 0.97 during the training and testing phases in the Lake Nokoué basin. Based on the evaluation indices used to assess the model’s performance for the appropriate forecast horizon of water level in the Lake Nokoué basin, the forecast horizon of t + 3 days is chosen for predicting future daily water levels.
Large Language Models (LLMs) combined with visual foundation models have demonstrated significant advancements, achieving intelligence levels comparable to human capabilities. This study analyzes the latest Multimodal LLMs (MLLMs), including Multimodal-GPT, GPT-4 Vision, Gemini, and LLaVa, with a focus on hydrological applications such as flood management, water level monitoring, agricultural water discharge, and water pollution management. We evaluated these MLLMs on hydrology-specific tasks, testing their response generation and real-time suitability in complex real-world scenarios. Prompts were designed to enhance the models’ visual inference capabilities and contextual comprehension from images. Our findings reveal that GPT-4 Vision demonstrated exceptional proficiency in interpreting visual data, providing accurate assessments of flood severity and water quality. Additionally, MLLMs showed potential in various hydrological applications, including drought prediction, streamflow forecasting, groundwater management, and wetland conservation. These models can optimize water resource management by predicting rainfall, evaporation rates, and soil moisture levels, thereby promoting sustainable agricultural practices. This research provides valuable insights into the potential applications of advanced AI models in addressing complex hydrological challenges and improving real-time decision-making in water resource management
Keywords Abstract Compression strength prediction Self-compacting concrete Artificial intelligence Concrete is one of the most common construction materials used all over the word. In estimating the strength properties of concrete, laboratory works need to be carried out. However, researchers have adopted predictive models in order to minimize the rigorous laboratory works in estimating the compressive strength and other properties of concrete. Self-compacting concrete which is an advanced form of construction is adopted mainly in areas where vibrations may not be possible due to complexity of the form work or reinforcement. This work is targeted at predicting the compressive strength of self-compacting concrete using artificial intelligence techniques. A comparative performance analysis of all techniques is presented. The outcomes demonstrated that training in a Deep Neural Network model with several hidden layers could enhance the performance of the suggested model. The artificial neural network (ANN) model, possesses a high degree of steadiness when compared to experimental results of concrete compressive strength. ANN was observed to be a strong predictive tool, as such is recommended for formulation of many civil engineering properties that requires predictions. Much time and resources are saved with artificial intelligence models as it eliminates the need for experimental test which sometimes delay construction works.
The prediction of hydrological flows (rainfall-depth or rainfall-discharge) is becoming increasingly important in the management of hydrological risks such as floods. In this study, the Long Short-Term Memory (LSTM) network, a state-of-the-art algorithm dedicated to time series, is applied to predict the daily water level of lake Nokoue in Benin. This paper aims to provide an effective and reliable method enable of reproducing the future daily water level of Lake Nokoue, which is influenced by a combination of two phenomena: rainfall and river flow (runoff from the Ouémé River, the Sô River, the Porto-Novo lagoon, and the Atlantic Ocean). Performance analysis based on the forecasting horizon indicates that LSTM can predict the water level of Nokoué Lake up to a forecast horizon of t+10 days. Performance metrics such as Root Mean Square Error (RMSE), coefficient of correlation (R²), Nash-Sutcliffe Efficiency (NSE), and Mean Absolute Error (MAE) agree on a forecast horizon of up to t+3 days. The values of these metrics remain stable for forecast horizons of t+1 days, t+2 days, and t+3 days. The values of R² and NSE are greater than 0.97 during the training and testing phases in the Nokoué Lake basin. Based on the evaluation indices used to assess the model's performance for the appropriate forecast horizon of water level in the Nokoué lake basin, the forecast horizon of t+3 days is chosen for predicting future daily water levels.
