Effective water management requires a large-scale understanding of agricultural irrigation systems and how they shift in response to various stressors. Here, we leveraged advances in Machine Learning and availability of very high resolution remote sensing imagery to help resolve this long-standing issue. To this end, we developed a deep learning model to classify irrigation systems at a regional scale using remote sensing imagery. After testing different model architectures, hyper parameters, class weights and image sizes, we selected a U-Net architecture with a Resnet-34 backbone for this purpose. We applied transfer learning to increase training efficiency and model performance. We considered four irrigation systems as well as urban and background areas as land use/cover classes, and applied the model to 8,600 very high resolution (1 m) images, labeled with ground-truth observations of irrigation types, in a case study in Idaho, USA. Images were obtained from the US Department of Agriculture’s National Agriculture Imagery Program. Our model achieved state-of-the-art performance for segmentation of different classes on the train data (85% to 94%), validation data (72% to 86%), and test data (70% to 86%), which attests to the efficacy of the model for the segmentation of images based on spatial features. Aside from leveraging deep learning and remote sensing for resolving the standing real-world problem of multiple irrigation type segmentation, this study develops and publicly shares labeled data, as well as a trained deep learning model, for irrigation type segmentation that can be applied/transferred to other regions globally. Furthermore, this study offers novel information about the impacts of transfer learning, imbalanced training data, and efficacy of various model structures for multiple irrigation type segmentation.

Sustainable urban surface runoff management is receiving increased attention due to environmental and ecological consequences related to urbanization. This study presents a useful new framework for green infrastructure (GI) planning for the management of urban runoff quality and quantity. The proposed framework considers physical, technical, economic, and multi-stakeholder aspects related to urban runoff management while simultaneously addressing the uncertainties of the decision-making process and model parameters. The methodology is applied to regionally locating and sizing low impact development-best management practices (LID-BMPs) while considering the hydrologic and hydraulic impacts. A decision-making multi-objective optimization framework is developed by integrating: 1) a Multi-Layer Perceptron neural network founded on a Storm Water Management Model (SWMM-MLP meta-model), 2) NSGA-II multi-objective optimization, 3) fuzzy a-cut technique, and 4) a decision-making support model based on social choice theory to elicit trade-offs among system cost and LID-BMP performance indicators. The decision-making model, based on Fuzzy Social Choice (FSC) theory, is applied to simulate consensus between stakeholders for a partially cooperative group decision-making problem. The proposed methodology is explored in a catchment located in the northeastern part of Tehran, Iran. Results showed that the SWMM could be effectively replaced with a MLP-based meta-model in simulation-optimization problems for urban runoff management. In the application of FSC methods, the optimal scenarios were effective in reducing the volume of urban runoff and contamination loads while maintaining the optimality of the operation costs. Considering the optimal LID scenario, a reduction of more than 99% in runoff volume and biochemical oxygen demand (BOD), and a decrease of more than 92% of total suspended solids (TSS), occurred for the lower bound of uncertainty (lower (left) end of the alpha-cut = 0.3). For the upper bound of uncertainty (upper (right) end of the alpha-cut level = 0.3), a maximum reduction of 57% was obtained. In applying FSC methods through the decision-making process, the Borda Counting method considered the preferences of all stakeholders best. Moreover, the proposed framework allows decision-makers to decide on the acceptability and reliability of the optimal management scenarios considering their preferences and uncertainties.

In this study, a new fuzzy methodology for a multi-objective optimization of reservoir Water Quality Monitoring Stations (WQMS) was developed, based on Transinformation Entropy (TE), the IRanian Water Quality Index (IRWQI), and fuzzy social choice considering uncertainties. The approach was utilized in the Karkheh Dam reservoir in Iran. The objective functions were: 1) minimizing costs, 2) minimizing redundant information and uncertainties, and 3) maximizing the spatial coverage of the network. A CE-QUAL-W2 model was used for the simulation of water quality variables. The IRWQI was computed to reveal a complete picture of the reservoir water quality. The TE quantities were calculated for each pair of potential stations. The TE values were plotted against the spatial distances among potential WQMS to obtain the TE–Distance (TE–D) curve, and minimize redundant information among stations, while providing coverage of the entire network. A multi-objective Genetic Algorithm (NSGA-II) was applied to obtain Pareto-optimal solutions taking stakeholder preference into account. The most preferred solution was then obtained using fuzzy social choice approaches to achieve a consensus. The fuzziness embedded in the decision-making procedure, the uncertainty in the value of mutual information, and the uncertainty in identifying the optimal distance among WQMS were also investigated. Results indicated that the three fuzzy social choice approaches (Borda Count, Minimax, and Approval Voting) led to the same number of optimized WQMS in each fuzzy alpha-cut. Based on the fuzzy linguistic quantifiers method, the number of optimized WQMS was increased.

A multi-objective optimization methodology is proposed herein for accurate identification of leakage in water distribution networks (WDNs) using pressure and flow sensors. We first model leakage at potential nodes using the EPANET software, and then divide WDN into near-homogenous zones using k-means clustering algorithm based on geographic distribution of nodes. Finally, flow and pressure sensors locations are optimized using the NSGA-II algorithm to identify the leakage zone accurately. Novelty of the proposed approach lies in sequential optimization of flow and pressure sensors placement, which helps improve the accuracy of leakage zone identification in WDNs. The objective functions of this study are: 1) maximizing accuracy of identified leakage zone and 2) minimizing number of sensors (and hence operational costs). Simulation results of the Mesopolis WDN corroborate the efficiency and effectiveness of the proposed approach.

Large volumes of water are wasted through leakage in water distribution networks, and early detection of leakages is important to minimize lost water. Pressure sensors can be placed in a network to detect changes in pressure that indicate the presence of a new leak. This study presents a new approach for placing a set of pressure sensors by creating a list of candidate locations based on sensitivity to leaks that are simulated at all potential nodes in a network. The selection of a set of sensors is explored for two objectives, which are the minimization of the number of sensors and the time of detection. The non-dominated sorting genetic algorithm (NSGA-II) is used to explore trade-offs between these objectives. The effect of measurement uncertainty on the selection of sensor locations is explored by identifying alternative non-dominated fronts for different values for sensor error. The evolutionary algorithm-based approach is applied and demonstrated for the C-Town water network.

Different biochars produced by the impregnation of Mg, Ca, Al, Cu, and Fe were compared for the phosphate (P) uptake capacity and the effect on solution pH. Among them, Ca- and Mg-rich biochars demonstrate better sorption ability to P and have less effect on pH change. The optimum conditions of the pyrolysis processes were determined using response surface methodology. Comparison of the P removal efficiency of these two biochars under optimum conditions imply the superior adsorption capability of Ca-rich biochar. According to XRD analysis, calcite is the dominant mineral on the biochar surface, indicating the potential of Ca-rich biochar for P removal by adsorption and precipitation. Predictive second-order kinetic and linear Langmuir isotherm models could adequately interpret the P sorption process for optimized Ca-rich biochar. The maximum P sorption capacity of Ca-rich biochar of 153.85 mg/g is superior to other adsorbents reported in literature.

Heatwaves are extended periods of unusually high temperatures with significant societal and environmental impacts. Despite their significance, there is not a generalized definition for heatwaves. In this paper, we introduce a multi-method global heatwave and warm-spell data record and analysis toolbox (named GHWR). In addition to a comprehensive long-term global data record of heatwaves, GHWR allows processing and extracting heatwave records for any location efficiently. We use traditional constant temperature threshold methods, as well as spatially and temporally localized threshold approaches to identify heatwaves. GHWR includes binary (0/1) occurrence records of heatwaves/warm-spells, and annual summary files with detailed information on their frequency, duration, magnitude and amplitude. GHWR also introduces the standardized heat index (SHI) as a generalized statistical metric to identify heatwave/warm-spells. SHI has direct association with the probability distribution function of long-term daily temperatures for any given calendar day and spatial grid. Finally, GHWR offers a unique opportunity for users to select the type of heatwave/warm-spell information from a plethora of methods based on their needs and applications.

In the present study, a BIOPLUME III simulation model is coupled with a non-dominating sorting genetic algorithm (NSGA-II)-based model for optimal design of in situ groundwater bioremediation system, considering preferences of stakeholders. Ministry of Energy (MOE), Department of Environment (DOE), and National Disaster Management Organization (NDMO) are three stakeholders in the groundwater bioremediation problem in Iran. Based on the preferences of these stakeholders, the multi-objective optimization model tries to minimize: (1) cost; (2) sum of contaminant concentrations that violate standard; (3) contaminant plume fragmentation. The NSGA-II multi-objective optimization method gives Pareto-optimal solutions. A compromised solution is determined using fallback bargaining with impasse to achieve a consensus among the stakeholders. In this study, two different approaches are investigated and compared based on two different domains for locations of injection and extraction wells. At the first approach, a limited number of predefined locations is considered according to previous similar studies. At the second approach, all possible points in study area are investigated to find optimal locations, arrangement, and flow rate of injection and extraction wells. Involvement of the stakeholders, investigating all possible points instead of a limited number of locations for wells, and minimizing the contaminant plume fragmentation during bioremediation are new innovations in this research. Besides, the simulation period is divided into smaller time intervals for more efficient optimization. Image processing toolbox in MATLAB((R)) software is utilized for calculation of the third objective function. In comparison with previous studies, cost is reduced using the proposed methodology. Dispersion of the contaminant plume is reduced in both presented approaches using the third objective function. Considering all possible points in the study area for determining the optimal locations of the wells in the second approach leads to more desirable results, i.e. decreasing the contaminant concentrations to a standard level and 20% to 40% cost reduction. (C) 2017 Elsevier B.V. All rights reserved.