Jan Adamowski

Comparison of Multivariate Regression and Artificial Neural Networks for Peak Urban Water-Demand Forecasting: Evaluation of Different ANN Learning Algorithms

For the past several years, Cyprus has been facing an unprecedented water crisis. Four options that have been considered to help resolve the problem of drought in Cyprus include imposing effective water use restrictions, implementing water-demand reduction programs, optimizing water supply systems, and developing sustainable alternative water source strategies. An important aspect of these initiatives is the accurate forecasting of short-term water demands, and in particular, peak water demands. This study compared multiple linear regression and three types of multilayer perceptron artificial neural networks (each of which used a different type of learning algorithm) as methods for peak weekly water-demand forecast modeling. The analysis was performed on 6 years of peak weekly water-demand data and meteorological variables (maximum weekly temperature and total weekly rainfall) for two different regions (Athalassa and Public Garden) in the city of Nicosia, Cyprus. 20 multiple linear regression models, 20 L...

Modeling of daily pan evaporation in sub tropical climates using ANN, LS-SVR, Fuzzy Logic, and ANFIS

Abstract This paper investigates the abilities of Artificial Neural Networks (ANN), Least Squares – Support Vector Regression (LS-SVR), Fuzzy Logic, and Adaptive Neuro-Fuzzy Inference System (ANFIS) techniques to improve the accuracy of daily pan evaporation estimation in sub-tropical climates. Meteorological data from the Karso watershed in India (consisting of 3801 daily records from the year 2000 to 2010) were used to develop and test the models for daily pan evaporation estimation. The measured meteorological variables include daily observations of rainfall, minimum and maximum air temperatures, minimum and maximum humidity, and sunshine hours. Prior to model development, the Gamma Test (GT) was used to derive estimates of the noise variance for each input–output set in order to identify the most useful predictors for use in the machine learning approaches used in this study. The ANN models consisted of feed forward backpropagation (FFBP) models with Bayesian Regularization (BR), along with the Levenberg–Marquardt (LM) algorithm. A comparison was made between the estimates provided by the ANN, LS-SVR, Fuzzy Logic, and ANFIS models. The empirical Hargreaves and Samani method (HGS), as well as the Stephens–Stewart (SS) method, were also considered for comparison with the newer machine learning methods. The Root Mean Square Error (RMSE) and Correlation Coefficient (CORR) were the statistical performance indices that were used to evaluate the accuracy of the various models. Based on the comparison, it was found that the Fuzzy Logic and LS-SVR approaches can be employed successfully in modeling the daily evaporation process from the available climatic data. In addition, results showed that the machine learning models outperform the traditional HGS and SS empirical methods.

Assessing the Impacts of Four Land Use Types on the Water Quality of Wetlands in Japan

This study examined how changes in the composition of land use can affect wetland water quality. Twenty-four wetlands located in Hiroshima prefecture in the western part of Japan were selected for this purpose. The water quality parameters that were explored include: pH, electrical conductivity, turbidity, dissolved oxygen, total dissolved solid, temperature and different forms of nitrogen. These important indicators of the water quality in the study area were measured from December 2005 to December 2006. The composition of land uses was determined for the catchments of the wetlands. They were then categorized into three classes, including non-disturbed, moderately-disturbed and highly-disturbed wetlands, based on the extent of urban area (as the most disruptive land use type within the catchment of the wetlands). The relationship between land use types and water quality parameters for the wetlands was statistically examined. The findings indicated that there were significant positive relationships between the proportion (%) of urban areas within catchments of the wetlands and EC (r = 0.67, p < 0.01), TDS (r = 0.69, p < 0.01), TN (r = 0.92, p < 0.01), DON (r = 0.6, p < 0.01), NH4+(r = 0.47, p < 0.05), NO2− (r = 0.50, p < 0.05), while negative relationships were observed between the proportion (%) of forest area in these wetlands and EC (r = −0.62, p < 0.01), TDS (r = −0.68, p < 0.01), TN (r = −0.68, p < 0.01), DON (r = -0.43, p < 0.05), and NH4+ (r = −0.55, p < 0.01). Analysis of the variance also revealed significant differences within the wetland groups in terms of the annual mean of electrical conductivity, total dissolved solids, total nitrogen, nitrite, dissolved inorganic nitrogen and dissolved organic nitrogen in the study area. Moreover, the study also indicated that the forest area plays a significant role in withholding nutrient loads from the wetlands, and hence, it can act as a sink for surface/subsurface nutrient inputs flowing into such water bodies from the watersheds.

Towards adaptive and integrated management paradigms to meet the challenges of water governance

Integrated Water Resource Management (IWRM) aims at finding practical and sustainable solutions to water resource issues. Research and practice have shown that innovative methods and tools are not sufficient to implement IWRM - the concept needs to also be integrated in prevailing management paradigms and institutions. Water governance science addresses this human dimension by focusing on the analysis of regulatory processes that influence the behavior of actors in water management systems. This paper proposes a new methodology for the integrated analysis of water resources management and governance systems in order to elicit and analyze case-specific management paradigms. It builds on the Management and Transition Framework (MTF) that allows for the examination of structures and processes underlying water management and governance. The new methodology presented in this paper combines participatory modeling and analysis of the governance system by using the MTF to investigate case-specific management paradigms. The linking of participatory modeling and research on complex management and governance systems allows for the transfer of knowledge between scientific, policy, engineering and local communities. In this way, the proposed methodology facilitates assessment and implementation of transformation processes towards IWRM that require also the adoption of adaptive management principles. A case study on flood management in the Tisza River Basin in Hungary is provided to illustrate the application of the proposed methodology.

Forecasting Urban Water Demand Via Wavelet-Denoising and Neural Network Models. Case Study: City of Syracuse, Italy

Forecasting urban water demand can be of use in the management of water utilities. For example, activities such as water-budgeting, operation and maintenance of pumps, wells, reservoirs, and mains require quantitative estimations of water resources at specified future dates. In this study, we tackle the problem of forecasting urban water demand by means of back-propagation artificial neural networks (ANNs) coupled with wavelet-denoising. In addition, non-coupled ANN and Linear Multiple Regression were used as comparison models. We considered the case of the municipality of Syracuse, Italy; for this purpose, we used a 7 year-long time series of water demand without additional predictors. Six forecasting horizons were considered, from 1 to 6 months ahead. The main objective was to implement a forecasting model that may be readily used for municipal water budgeting. An additional objective was to explore the impact of wavelet-denoising on ANN generalization. For this purpose, we measured the impact of five different wavelet filter-banks (namely, Haar and Daubechies of type db2, db3, db4, and db5) on a single neural network. Empirical results show that neural networks coupled with Haar and Daubechies’ filter-banks of type db2 and db3 outperformed all of the following: non-coupled ANN, Multiple Linear Regression and ANN models coupled with Daubechies filters of type db4 and db5. The results of this study suggest that reduced variance in the training-set (by means of denoising) may improve forecasting accuracy; on the other hand, an oversimplification of the input-matrix may deteriorate forecasting accuracy and induce network instability.

Standard precipitation index drought forecasting using neural networks, wavelet neural networks, and support vector regression

Drought forecasts can be an effective tool for mitigating some of the more adverse consequences of drought. Data-driven models are suitable forecasting tools due to their rapid development times, as well as minimal information requirements compared to the information required for physically based models. This study compares the effectiveness of three data-driven models for forecasting drought conditions in the Awash River Basin of Ethiopia. The Standard Precipitation Index (SPI) is forecast and compared using artificial neural networks (ANNs), support vector regression (SVR), and wavelet neural networks (WN). SPI 3 and SPI 12 were the SPI values that were forecasted. These SPI values were forecast over lead times of 1 and 6 months. The performance of all the models was compared using RMSE, MAE, and R2. The forecast results indicate that the coupled wavelet neural network (WN) models were the best models for forecasting SPI values over multiple lead times in the Awash River Basin in Ethiopia.

Empowering marginalized communities in water resources management: Addressing inequitable practices in Participatory Model Building

Within the field of water resource management, Group Model Building (GMB) is a growing method used to engage stakeholders in the development of models that describe environmental and socioeconomic systems to create and test policy alternatives. While there is significant focus on improving stakeholder engagement, there is a lack of studies specifically looking at the experiences of marginalized communities and the barriers that prevent their fuller participation in the decision-making process. This paper explores the common issues and presents recommended improved practices, based on anti-oppression, related to the stages of problem framing, stakeholder identification and selection, workshop preparation, and workshop facilitation. For problem defining and stakeholder selection, the major recommendations are to engage diverse stakeholder communities from the earliest stages and give them control over framing the project scope. With regards to planning the model building workshops, it is recommended that the facilitation team work closely with marginalized stakeholders to highlight and address barriers that would prevent their inclusion. With the actual facilitation of the workshops, it is best to employ activities that allow stakeholders to provide knowledge and input in mediums that are most comfortable to them; additionally, the facilitation team needs to be able to challenge problematic interpersonal interactions as they manifest within conversations. This article focuses on building comfortability with political language so that the systemic oppression in which existing participatory processes occur can be understood, thus allowing GMB practitioners to engage in social justice efforts.

Using causal loop diagrams for the initialization of stakeholder engagement in soil salinity management in agricultural watersheds in developing countries: A case study in the Rechna Doab watershed, Pakistan

Over the course of the last twenty years, participatory modeling has increasingly been advocated as an integral component of integrated, adaptive, and collaborative water resources management. However, issues of high cost, time, and expertise are significant hurdles to the widespread adoption of participatory modeling in many developing countries. In this study, a step-wise method to initialize the involvement of key stakeholders in the development of qualitative system dynamics models (i.e. causal loop diagrams) is presented. The proposed approach is designed to overcome the challenges of low expertise, time and financial resources that have hampered previous participatory modeling efforts in developing countries. The methodological framework was applied in a case study of soil salinity management in the Rechna Doab region of Pakistan, with a focus on the application of qualitative modeling through stakeholder-built causal loop diagrams to address soil salinity problems in the basin. Individual causal loop diagrams were developed by key stakeholder groups, following which an overall group causal loop diagram of the entire system was built based on the individual causal loop diagrams to form a holistic qualitative model of the whole system. The case study demonstrates the usefulness of the proposed approach, based on using causal loop diagrams in initiating stakeholder involvement in the participatory model building process. In addition, the results point to social-economic aspects of soil salinity that have not been considered by other modeling studies to date.

A novel multi criteria decision making model for optimizing time-cost-quality trade-off problems in construction projects

Using evidential reasoning as the multi-criterion decision making (MCDM) approach.A comprehensive framework to integrate MCDM methods with optimization techniques.Using MOGA with NSGA-II to solve discrete time-cost-quality trade-off problem.Determining the weights of the objectives with Shannons Entropy technique.Obtaining all the Pareto solutions for the 18-activity network benchmark example. The planning phase of every construction project is entangled with multiple and occasionally conflicting criteria which need to be optimized simultaneously. Multi-criterion decision-making (MCDM) approaches can aid decision-makers in selecting the most appropriate solution among numerous potential Pareto optimal solutions. An evidential reasoning (ER) approach was applied for the first time in the context of project scheduling to identify the best Pareto solution for discrete time-cost-quality trade-off problems (DTCQTPs). An exhaustive framework to synthesize the MCDM approaches with multi-objective optimization techniques was also proposed. To identify all global Pareto optimal solutions, a multi-objective genetic algorithm (MOGA) incorporating the NSGA-II procedure was developed and tested in a highway construction project case study. The Shannons entropy technique served to determine the relative weights of the objectives according to their contributions to the uncertainty of the results obtained. A benchmark case study of DTCQTP was solved using the proposed methodology, and the Pareto optimal solutions obtained were subsequently ranked using the ER approach. By investigating the performance of each scheduling alternative based on multiple criteria (e.g., time, cost, and quality), the proposed approach proved effective in raising the efficiently of construction project scheduling.

Exploring the behavioural attributes, strategies and contextual knowledge of champions of change in the Canadian water sector

Sustainable water resource management (WRM) is failing to be fully implemented in Canada due to, among other things, cultural and structural inhibiting factors. There is a need for water professionals to develop their understanding of the ways in which cultural and structural barriers within prominent water resource management institutions can be broken down and/or navigated so that climate change and sustainability challenges can be more appropriately addressed. This study explored, for the first time in Canada, champion leadership approaches by interviewing champions in the Canadian water sector, with a focus on behavioural attributes, strategies and contextual factors. The findings revealed the significance of both formal and informal relationships, passion in communication, respectful and humble networking and work relations alongside necessary risk taking as key behavioural strategies for Canadian water champions. It also exposed the need to understand contextual realities of mandate gaps, control and secrecy at the federal level versus the more open and responsive culture at the municipal level. While the context can inhibit change, it does not necessarily inhibit it if the champion is well equipped to understand the institution and the strategies that can influence it. Such strategies include the creation of windows of opportunities and the use of media such as journalists, for risk-taking change efforts that do not have to be socially and professionally threatening. Water professionals who have a better understanding of the champion experience in Canada may be in a better position to contribute to a more effective implementation of sustainable WRM in Canada.