Water Demand Forecasting Research Articles

Forecasting water consumption on transboundary water resources for water resource management using the feed-forward neural network: a case study of the Nile River in Egypt and Kenya

Water resources are an essential component of a country’s natural resource potential. Pressure on these resources is set to increase due to increased water demand, climate change and rainfall variability. This could lead to conflicts between sectoral users, within or between countries, especially among transboundary countries. Interest in transboundary water resources is a priority, especially where issues such as uncertainty regarding the status of transboundary waterbodies and reductions in water volume persist. In this study, we used the feed-forward neural network to forecast water demand along the Nile River in two countries, Egypt and Kenya. Two scenarios were modelled. Input data for the first scenario included preceding records of precipitation, gross domestic product, population and water use in the agricultural sector. The second scenario observed the effects of the growing economy on water resources by doubling the gross domestic product and keeping all other inputs constant. For Kenya, the results projected a steady increase in water demand throughout the next 20 years for both scenarios. However, for Egypt, the observed trend in both scenarios was a decline in water demand, followed by a steady increase. The results underscore the importance of forecasting for easier future planning and management, and to help governing bodies along transboundary water resources develop timely strategies in the future to alleviate future water shortages and poor management of water resources.

Urban Water Demand: Statistical Optimization Approach to Modeling Daily Demand

AbstractReliable forecasts of water demand that account for factors that drive demand are imperative to understanding future urban water needs. The effects of meteorological dynamics and sociocultu...

The Drought Resilience Impact Platform (DRIP): Improving Water Security Through Actionable Water Management Insights

Millions of people living in the drought prone Horn of Africa face an increasing threat from a lack of safe, reliable and affordable water year-round as droughts become more severe and frequent. Drought emergencies emerge when reduced rainfall conspires with limited community water service capacity to cause dramatic reductions in access to water for people, livestock and agriculture. Drought-driven humanitarian emergencies can be prevented if groundwater is reliably made available at strategic locations during cycles of water stress. DRIP - The Drought Resilience Impact Platform, is an initiative combining early detection and planning with proactive groundwater management to ensure water availability, thus enabling drought-prone communities to become effective managers in the prevention of these humanitarian crises. It replaces reactive and expensive short-term assistance measures, like water trucking, with a framework for drought resilience. DRIP links in-situ sensors deployed in East Africa with remote sensing data to improve estimates for rainfall and groundwater availability, and will develop a localized model for water demand forecasting. These indicators support the operation and maintenance of strategically selected groundwater borehole systems, thereby helping to support water delivery during dry and drought seasons. DRIP can be used to support pay-for-performance contracting, ensuring that water asset management is incentivized. DRIP is presently monitoring the water supplies of about 3 million people in East Africa. This paper presents DRIP's current web-based functionality, which uses several custom and commercial tools, and DRIP's applications, including rainfall-adjusted indicators of water pump functionality in Kenya and Ethiopia. Future work includes experimental and statistical characterization of the impact of these capabilities on water security, and the development of forecasting capabilities. This work is supported NASA, USAID, and the National Science Foundation. The views expressed in this article do not necessarily reflect the views of the United States Agency for International Development or the United States Government.

Chance-Constrained Water Pumping to Manage Water and Power Demand Uncertainty in Distribution Networks

Water pumping in drinking water distribution networks (WDNs) can be treated as a flexible load in the power distribution network (PDN). In this article, we formulate an optimization problem to minimize the electricity costs associated with pumping subject to WDN and PDN constraints. In practice, both water and power demands are uncertain and pumps should be scheduled to ensure that pump operation does not violate either networks’ constraints for nearly all possible uncertainty realizations. To address this problem, we formulate a chance-constrained (CC) optimization problem that simultaneously determines pumping schedules along with the parameters of real-time control policies that can be used to respond to water and power demand forecast errors. We use approximations and relaxations along with the scenario approach for CC programming to reformulate the optimization problem into a convex deterministic problem. We demonstrate the performance of the approach through case studies and also explore the impact of the relaxations, an approach to improve computational tractability, and tradeoffs associated with the way in which we define the cost of real-time control actions. We find that optimal scheduling and real-time control of water pumping can effectively manage water and power demand uncertainty, meaning water demand is satisfied and both the WDN and PDN operate within their limits; however, the approach is conservative leading to high reliability at high cost.

Analysis of driving forces of water demand in Jing-Jin-Ji district in recent years based on water demand field theory

In recent years, urbanization has accelerated and the gap between urban water supply and demand has become more significant. This phenomenon has led to greater demands on urban water resource management. Thus, analyses of the forces that drive urban water demand growth can help efficient management of urban water resources. “Jing-Jin-Ji” refers to the district composed of Beijing and Tianjin, municipality as well as Hebei Province, which is among the political and economic centers of China. The traditional analysis of driving force for water resources often consider cities as independent individuals and ignores the interaction between them. Such method cannot describe the spatial distribution of regional water demand. In this study, the water demand field theory is used to construct water demand and social development fields for the Jing-Jin-Ji district with relevant indicators. The distribution trends of the two fields, and the correlation of their intensities, are then analyzed. Principal component analysis is used to specifically determine the forces that drive water demand at different stages of development. Results indicate that the water demand field theory can accurately describe spatial variations in water demand intensity in the study area. The main driving force for water demand in the Jing-Jin-Ji district from 2000 to 2014 is the growth of urban population. However, in 2015 and 2016, “GDP of the tertiary industry” overtook it. The results of this study can serve as a basis for regional water demand forecasting, with large potential to be applied to urban water resource management.

Forecasting of monthly stochastic signal of urban water demand: Baghdad as a case study

Forecasting of municipal water demand is essential for the decision-making process in the water industry in particular for countries that suffered from water scarcity. An accurate prediction of water demand improves the water distribution systems’ performance. This study analyses the water consumption data of Baghdad city using a signal pre-treatment processing approach aiming at a stochastic signal extraction of such data. An autoregressive (AR) model is then applied to predict monthly water consumption. Our prediction model has been trained and tested using a water consumption data captured from Al-Wehda treatment plant between 2006 and 2015. The results reveal that applying signal pre-treatment method was an effective approach for detecting stochastics of our water consumption data, and the hybrid model was reliable for the prediction of water demand.

Hourly and Daily Urban Water Demand Predictions Using a Long Short-Term Memory Based Model

This case study uses a long short-term memory (LSTM)-based model to predict short-term urban water demands for the Hefei City of China. The performance of the LSTM-based model is compared with the autoregressive integrated moving average (ARIMA) model, the support vector regression (SVR) model, and the random forests (RF) model based on data with time resolutions ranging from 15 min to 24 h. Additionally, this paper investigates the performance of the LSTM-based model in predicting multiple successive data points. Results show that the LSTM-based model can offer predictions with improved accuracy than the other models when dealing with data with high time resolutions, data points with abrupt changes, and data with a relatively high uncertainty level. It is also observed that the LSTM-based model exhibits the best performance in predicting multiple successive water demands with high time resolutions. In addition, the inclusion of external parameters (e.g., temperature) cannot enhance the performance of the LSTM-based model, but it can improve ARIMAX's prediction ability (ARIMAX is the ARIMA with variables). These observations provide additional and improved evaluations regarding the LSTM-based models used for short-term urban water demand forecasting, thereby enabling their wider adoption in practical applications.

Hybrid Model for Short-Term Water Demand Forecasting Based on Error Correction Using Chaotic Time Series

Short-term water demand forecasting plays an important role in smart management and real-time simulation of water distribution systems (WDSs). This paper proposes a hybrid model for the short-term forecasting in the horizon of one day with 15 min time steps, which improves the forecasting accuracy by adding an error correction module to the initial forecasting model. The initial forecasting model is firstly established based on the least square support vector machine (LSSVM), the errors time series obtained by comparing the observed values and the initial forecasted values is next transformed into chaotic time series, and then the error correction model is established by the LSSVM method to forecast errors at the next time step. The hybrid model is tested on three real-world district metering areas (DMAs) in Beijing, China, with different demand patterns. The results show that, with the help of the error correction module, the hybrid model reduced the mean absolute percentage error (MAPE) of forecasted demand from (5.64%, 4.06%, 5.84%) to (4.84%, 3.15%, 3.47%) for the three DMAs, compared with using LSSVM without error correction. Therefore, the proposed hybrid model provides a better solution for short-term water demand forecasting on the tested cases.

Short-term water demand forecasting using hybrid supervised and unsupervised machine learning model

Regression Tree (RT) forecasting models are widely used in short-term demand forecasting. Likewise, Self-Organizing Maps (SOM) models are known for their ability to cluster and organize unlabeled big data. Herein, a combination of these two Machine Learning (ML) techniques is proposed and compared to a standalone RT and a Seasonal Autoregressive Integrated Moving Average (SARIMA) models, in forecasting the short-term water demand of a municipality. The inclusion of the Unsupervised Machine Learning clustering model has resulted in a significant improvement in the performance of the Supervised Machine Learning forecasting model. The results show that using the output of the SOM clustering model as an input for the RT forecasting model can, on average, double the accuracy of water demand forecasting. The Mean Absolute Percentage Error (MAPE) and the Normalized Root Mean Squared Error (NRMSE) were calculated for the proposed models forecasting 1 h, 8 h, 24 h, and 7 days ahead. The results show that the hybrid models outperformed the standalone RT model, and the broadly used SARIMA model. On average, hybrid models achieved double accuracy in all 4 forecast periodicities. The increase in forecasting accuracy afforded by this hybridized modeling approach is encouraging. In our application, it shows promises for more efficient energy and water management at the water utilities.

Forecasting water demand under climate change using artificial neural network: a case study of Kathmandu Valley, Nepal

Abstract With a water demand of 370 MLD, Kathmandu Valley is currently facing a water shortage of 260 MLD. The Melamchi Water Supply Project (MWSP) is an interbasin project aimed at diverting 510 MLD of water in three phases (170 MLD in each phase). Phase I of the project was expected to complete by 2018. Water demand forecasting is the first and important activity in managing water supply. Using the socio-economic factors of number of connections, water tariff and ratio of population to number of university students and climatic factor of annual rainfall, artificial neural network (ANN) was used to predict the water demand of Kathmandu Valley until the year 2040. The analysis suggests that, even after the completion of Phase I of MWSP, the water scarcity in the valley will be 160 MLD in 2020. Therefore, Phase II of MWSP should be completed within 2025 and Phase III should be completed by 2040. The result of this study aids KUKL for better management of the water system. In addition, this research can help in decision making to construct the second and third phase for MWSP, the construction date of which still has not been decided.

Forecasting water demand across a rapidly urbanizing region

Urban growth and climate change together complicate planning efforts meant to adapt to increasingly scarce water supplies. Several studies have independently examined the impacts of urban planning and climate change on water demand, but little attention has been given to their combined impact. Here we forecast urban water demand using a Geographically Weighted Regression model informed by socio-economic, environmental and landscape pattern metrics. The purpose of our study is to evaluate how future scenarios of population densities and climate warming will jointly affect water demand across two rapidly growing U.S. states (North Carolina and South Carolina). Our forecasts indicate that regional water demand by 2065 will increase by 37%–383% relative to the baseline in 2010, across all scenarios of change. Our results show future water demand will increase under rising temperatures, but could be ameliorated by policies that promote higher density development and urban infill. These water-efficient land use policies show a 5% regional reduction in water demand and up to 25% reduction locally for counties with the highest expected population growth by 2065. For rural counties experiencing depopulation, the land use policies we considered are insufficient to significantly reduce water demand. For expanding communities seeking to increase their adaptive capacity to changing socio-environmental conditions, our framework can assist in developing sustainable solutions.

An ensemble stacked model with bias correction for improved water demand forecasting

ABSTRACT Water demand forecasting is an essential task for water utilities, with increasing importance due to future societal and environmental changes. This paper suggests a new methodology for water demand forecasting, based on model stacking and bias correction that predicts daily demands for groups of ~120 properties. This methodology is compared to a number of models (Artificial Neural Networks – ANNs, Generalised Linear Models – GLMs, Random Forests – RFs, Gradient Boosting Machines – GBMs, Extreme Gradient Boosting – XGBoost, and Deep Neural Networks – DNNs), using real consumption data from the UK, collected at 15–30 minute intervals from 1,793 properties. Results show that the newly proposed stacked model that comprises of RFs, GBMs, DNNs, and GLMs consistently outperformed other water demand forecasting techniques (peak R2 = 74.1%). The stacked model’s accuracy on peak consumption days further improved by applying a bias correction method on the model’s output.

Analysis of Current and Future WaterYejube is one of the rapidly developing towns in Ethiopia. Difficulties of acceptable water supplies and reduction of non-revenue water remains one of the main challenges facing Demand Scenario in Yejube Town, Ethiopia

Yejube is one of the rapidly developing towns in Ethiopia. Difficulties of acceptable water supplies and reduction of non-revenue water remains one of the main challenges facing Yejube water supply system. This study is concentration on the analysis of water demand scenario and future water demand forecasting of for Yejube town. A mixed cross-sectional research method was implemented to analyzed the past and future water demand of the town. The outcome of this study found that the municipal average daily water consumption per capita was 14.3l/c/day and 16l/c/day respectively in 2017 and 2018. The water supply coverage of the town is low when we compare the regional and international standards. Water production and demand are unbalanced, due to continuous interruption of the water supply system inhabitants are facing water related problems. And also, 20.38% and 24% of the water production lost in the distribution system before reaching to the end customers. The population of the town in 2040 expected to reach 24807 with an average daily per capita water consumption of 36.7l/day. Generally, delivery of hygienic Water Supply is one of the key influencing factors that significantly subsidize to the socioeconomic renovation of a country by enlightening the health thereby increasing life standard and economic productivity of the society. Therefore, accurate forecast of future water demand plays a crucial role in optimizing the need for potable water resources and effective water allocation among competing users. Keywords : forecasting, unaccounted for water, water demand forecast, water demand, Yejube town, DOI: 10.7176/CER/12-3-02 Publication date: March 31 st 2020

A Method for Predicting Long-Term Municipal Water Demands Under Climate Change

The accurate forecast of water demand is challenging for water utilities, specifically when considering the implications of climate change. As such, this is the first study that focuses on finding associations between monthly climate factors and municipal water consumption, using baseline data collected between 1980 and 2010. The aim of the study was to investigate the reliability and capability of a combination of techniques, including Singular Spectrum Analysis (SSA) and Artificial Neural Networks (ANNs), to accurately predict long-term, monthly water demands. The principal findings of this research are as follows: a) SSA is a powerful method when applied to remove the impact of socio-economic variables and noise, and to determine a stochastic signal for long-term water consumption time series; b) ANN performed better when optimised using the Lightning Search Algorithm (LSA-ANN) compared with other approaches used in previous studies, i.e. hybrid Particle Swarm Optimisation (PSO-ANN) and Gravitational Search Algorithm (GSA-ANN); c) the proposed LSA-ANN methodology was able to produce a highly accurate and robust model of water demand, achieving a correlation coefficient of 0.96 between observed and predicted water demand when using a validation dataset, and a very small root mean square error of 0.025.

Machine Learning and Data Analytic Techniques in Digital Water Metering: A Review

Digital or intelligent water meters are being rolled out globally as a crucial component in improving urban water management. This is because of their ability to frequently send water consumption information electronically and later utilise the information to generate insights or provide feedback to consumers. Recent advances in machine learning (ML) and data analytic (DA) technologies have provided the opportunity to more effectively utilise the vast amount of data generated by these meters. Several studies have been conducted to promote water conservation by analysing the data generated by digital meters and providing feedback to consumers and water utilities. The purpose of this review was to inform scholars and practitioners about the contributions and limitations of ML and DA techniques by critically analysing the relevant literature. We categorised studies into five main themes: (1) water demand forecasting; (2) socioeconomic analysis; (3) behaviour analysis; (4) water event categorisation; and (5) water-use feedback. The review identified significant research gaps in terms of the adoption of advanced ML and DA techniques, which could potentially lead to water savings and more efficient demand management. We concluded that further investigations are required into highly personalised feedback systems, such as recommender systems, to promote water-conscious behaviour. In addition, advanced data management solutions, effective user profiles, and the clustering of consumers based on their profiles require more attention to promote water-conscious behaviours.

Smart meters data for modeling and forecasting water demand at the user-level

Smart meters installed at the user-level provide a new data source for managing water infrastructure. This research explores the use of machine learning methods, including Random Forests (RFs), Artificial Neural Networks (ANNs), and Support Vector Regression (SVR) to forecast hourly water demand at 90 accounts using smart-metered data. Demands are predicted using lagged demand, seasonality, weather, and household characteristics. Time-series clustering is applied to delineate data based on the time of day and day of the week, which improves model performance. Two modeling approaches are compared. Individual models are developed separately for each meter, and a Group model is trained using a data set of multiple meters. Individual models predict demands at meters in the original data set with lower error than Group models, while the Group model predicts demands at new meters with lower error than Individual models. Results demonstrate that RF and ANN perform better than SVR across all scenarios.

State-of-art of genetic programming applications in water-resources systems analysis.

Evolutionary algorithms (EAs) have become competitive solvers of a wide variety of water-resources optimization problems. Genetic programming (GP) has become a leading EA since its inception in 1985. This paper reviews the state-of-the-art of GP and its applications in water-resources systems analysis. A comprehensive knowledge about GP's theory and modeling approach is essential for its successful application in water-resources systems analysis. This review presents variants of GP that have been proven useful in various applications to water resources problems. Several examples of applications of GP in water-resources systems analysis are herein presented. This review reveals GP's capability and superiority compared to other conventional methods, which makes it suitable for solving a wide variety of water-related problems including rainfall-runoff modeling, streamflow sediment prediction, flood prediction and routing, evaporation and evapotranspiration forecasting, reservoir operation, groundwater modeling, water quality modeling, water demand forecasting, and water distribution systems.

Applying human mobility and water consumption data for short-term water demand forecasting using classical and machine learning models

ABSTRACTWater demand forecasting is a crucial task in the efficient management of the water supply system. This paper compares classical and adapted machine learning algorithms used for water usage predictions including ARIMA, support vector regression, random forests and extremely randomized trees. These models were enriched with human mobility data to improve the predictive power of water demand forecasting. Furthermore, a framework for processing mobility data into time-series correlated with water usage data is proposed. This study uses 51 days of water consumption readings and over 7 million geolocated mobility records from urban areas. Results show that using human mobility data improves water demand prediction. The best forecasting algorithm employing a random forest method achieved 90.4% accuracy (measured by the mean absolute percentage error) and is better by 1% than the same algorithm using only water data, while classic ARIMA approach achieved 90.0%. The Blind (copying) prediction achieved 85.1% of accuracy.

A Two-Layer Water Demand Prediction System in Urban Areas Based on Micro-Services and LSTM Neural Networks

In recent years, scarce water resources became one of the main problems that endanger human species existence and the advancement of any nation. In this research, smart water meters were implemented, distributed, and installed in a regional area in Cairo while data were collected at uniform intervals then sent to the cloud instantly. The solution paradigm uses an Internet of Things (IoT) based on micro-services and containers. The design incorporates real-time streaming and infrastructure performance optimization to store data. A second layer to analyze the acquired data was used to model water consumption using Long Short-Term Memory (LSTM). The designed LSTM is validated and tested to be utilized in the forecast of future water demand. Moreover, two alternative machine learning methods, namely Support Vector Regression and Random Forest commonly utilized in time series forecasting applications, were used for a comparative analysis of which LSTM has proven to be superior. The proper integration of the system elements is the key to the proposed system success. Based on the success of the designed system, it can be applicable on a national scale. That can enable the optimal management of consumers' demand and improve water infrastructure utilization. The proposed paradigm presents a testbed for various scenarios that can be used in water resources management.

Prediction and Analysis of Water Resources using Machine Learning Algorithm

Water demand prediction plays an important role in urban and environmental planning, ecological development, decision-making processes and optimum utilization of water resources. A precise water demand prediction has a key job in the forecasting, design, process, and organisation of water resources frameworks. The under stress natural resources and the ever increasing population size makes it dominant to accurately and efficiently forecast water demand in the urban area which is possible by applying data mining techniques on the huge volumes of available water data. This paper focuses on building precise predictive models for water demand prediction using support vector machine which takes care of the nonlinear changeability of water demand at diverse levels for optimal operations.