OPTIMIZATION OF THE HYDROMETRIC NETWORK OPERATION BY A HEURISTIC TRAVELING SALESMAN ALGORITHM1

ABSTRACT: The problem of determining the optimal route to operate a network of hydrometric gauges within a given area is formulated and solved by an original heuristic traveling salesman algorithm. The algorithm used for solving the problem takes into account the variation of station service time, the eight-hour maximum working day, existence of more than one home base, and the budget constraint. The traveling salesman algorithm is used to solve the optima! assignment problem within a one-day time limit, locating a starting base, stations to visit, and the closest home base. The ending base for the previous day becomes the starting base for the next day, and the assignment procedure is repeated. The computation proceeds from day to day, until all the stations in the considered region are serviced. The solution obtained presents the optimal route to operate a network of hydrometric gauges. The optimization procedure has been applied to the Dauphin hydrometric field area in Manitoba.     

HYDROLOGIC UNCERTAINTY MEASURE AND NETWORK DESIGN1

ABSTRACT: This paper presents a simple methodology, using the entropy concept, to estimate regional hydro logic uncertainty and information at both gaged and ungaged grids in a basin. The methodology described in this paper is applicable for (a) the selection of the optimum station from a dense network, using maximization of information transmission criteria, and (b) expansion of a network using data from an existing sparse network by means of the information interpolation concept and identification of the zones from minimum hydrologic information. The computation of single and joint entropy terms used in the above two cases depends upon single and multivariable probability density functions. In this paper, these terms are derived for the gamma distribution. The derived formulation for optimum hydrologic network design was tested using the data from a network of 29 rain gages on Sleeper River Experimental Watershed. For the purpose of network reduction, the watershed was divided into three subregions, and the optimum stations and their locations in each subregion were identified. To apply the network expansion methodology, only the network consisting of 13 stations was used, and feasible triangular elements were formed by joining the stations. Hydrologic information was calculated at various points on the line segments, and critical information zones were identified by plotting information contours. The entropy concept used in this paper, although derived for single and bivaviate gamma distribution, is general in type and can easily be modified for other distributions by a simple variable transformation criterion.     

SENSITIVITY ANALYSIS,CALIBRATION, AND VALIDATIONS FOR A MULTISITE AND MULTIVARIABLE SWAT MODEL1

The ability of a watershed model to mimic specified watershed processes is assessed through the calibration and validation process. The Soil and Water Assessment Tool (SWAT) watershed model was implemented in the Beaver Reservoir Watershed of Northwest Arkansas. The objectives were to: (1) provide detailed information on calibrating and applying a multisite and multivariable SWAT model; (2) conduct sensitivity analysis; and (3) perform calibration and validation at three different sites for flow, sediment, total phosphorus (TP), and nitrate‐nitrogen (NO₃‐N) plus nitrite‐nitrogen (NO₂‐N). Relative sensitivity analysis was conducted to identify parameters that most influenced predicted flow, sediment, and nutrient model outputs. A multi objective function was defined that consisted of optimizing three statistics: percent relative error (RE), Nash‐Sutcliffe Coefficient (R_NS²), and coefficient of determination (R²). This function was used to successfully calibrate and validate a SWAT model of Beaver Reservoir Watershed at multi‐sites while considering multivariables. Calibration and validation of the model is a key factor in reducing uncertainty and increasing user confidence in its predictive abilities, which makes the application of the model effective. Information on calibration and validation of multisite, multivariable SWAT models has been provided to assist watershed modelers in developing their models to achieve watershed management goals.     

Assessing Linkages in Stream Habitat,Geomorphic Condition,and Biological Integrity Using a Generalized Regression Neural Network

Watershed managers often use physical geomorphic and habitat assessments in making decisions about the biological integrity of a stream, and to reduce the cost and time for identifying stream stressors and developing mitigation strategies. Such analysis is difficult since the complex linkages between reach‐scale geomorphic and habitat conditions, and biological integrity are not fully understood. We evaluate the effectiveness of a generalized regression neural network (GRNN) to predict biological integrity using physical (i.e., geomorphic and habitat) stream‐reach assessment data. The method is first tested using geomorphic assessments to predict habitat condition for 1,292 stream reaches from the Vermont Agency of Natural Resources. The GRNN methodology outperforms linear regression (69% vs. 40% classified correctly) and improves slightly (70% correct) with additional data on channel evolution. Analysis of a subset of the reaches where physical assessments are used to predict biological integrity shows no significant linear correlation, however the GRNN predicted 48% of the fish health data and 23% of macroinvertebrate health. Although the GRNN is superior to linear regression, these results show linking physical and biological health remains challenging. Reasons for lack of agreement, including spatial and temporal scale differences, are discussed. We show the GRNN to be a data‐driven tool that can assist watershed managers with large quantities of complex, nonlinear data.     

Effectiveness of Multi‐Site Weather Generator for Hydrological Modeling1

Khalili, Malika, François Brissette, and Robert Leconte, 2011. Effectiveness of Multi‐site Weather Generator for Hydrological Modeling. Journal of the American Water Resources Association (JAWRA) 1‐12. DOI: 10.1111/j.1752‐1688.2010.00514.x Abstract: A multi‐site weather generator has been developed using the concept of spatial autocorrelation. The multi‐site generation approach reproduces the spatial autocorrelations observed between a set of weather stations as well as the correlations between each pair of stations. Its performance has been assessed in two previous studies using both precipitation and temperature data. The main objective of this paper is to assess the efficiency of this multi‐site weather generator compared to a uni‐site generator with respect to hydrological modeling. A hydrological model, known as Hydrotel, was applied over the Chute du Diable watershed, located in the Canadian province of Quebec. The distributed nature of Hydrotel accounts for the spatial variations throughout the watershed, and thus allows a more in‐depth assessment of the effect of spatially dependent meteorological input on runoff generation. Simulated streamflows using both the multi‐site and uni‐site generated weather data were statistically compared to flows modeled using observed data. Overall, the hydrological modeling using the multi‐site weather generator significantly outperformed that using the uni‐site generator. This latter combined to Hydrotel resulted in a significant underestimation of extreme streamflows in all seasons.     

A heuristic optimization approach for Air Quality Monitoring Network design with the simultaneous consideration of multiple pollutants

An interactive optimization methodology for allocating the number and configuration of an Air Quality Monitoring Network (AQMN) in a vast area to identify the impact of multiple pollutants is described. A mathematical model based on the multiple cell approach (MCA) was used to create monthly spatial distributions for the concentrations of the pollutants emitted from different emission sources. These spatial temporal patterns were subject to a heuristic optimization algorithm to identify the optimal configuration of a monitoring network. The objective of the optimization is to provide maximum information about multi-pollutants (i.e., CO, NO(x) and SO(2)) emitted from each source within a given area. The model was applied to a network of existing refinery stacks and the results indicate that three stations can provide a total coverage of more than 70%. In addition, the effect of the spatial correlation coefficient (R(C)) on total area coverage was analyzed. The modeling results show that as the cutoff correlation coefficient R(C) is increased from 0.75 to 0.95, the number of monitoring stations required for total coverage is increased. A high R(C) based network may not necessarily cover the entire region, but the covered region will be well represented. A low R(C) based network, on the other hand, would offer more coverage of the region, but the covered region may not be satisfactorily represented.     

Locating Existing Best Management Practices Within a Watershed: The Value of Multiple Methods

There is an increasing need to document the impacts of conservation‐related best management practices (BMPs) on water quality within a watershed. However, this impact analysis depends upon accurate geospatial locations of existing practices, which are difficult to obtain. This study demonstrates and evaluates three different methods for obtaining geospatial information for BMPs. This study was focused on the Eagle Creek Watershed, a mixed use watershed in central Indiana. We obtained geospatial information for BMPs through government records, producer interviews, and remote‐sensing aerial photo interpretation. Aerial photos were also used to validate the government records and producer interviews. This study shows the variation in results obtained from the three sources of information as well as the benefits and drawbacks of each method. Using only one method for obtaining BMP information can be incomplete, and this study demonstrates how multiple methods can be used for the most accurate picture.     

SCREENING THE RELOCATION STRATEGIES OF WATER QUALITY MONITORING STATIONS BY COMPROMISE PROGRAMMING1

Water quality monitoring network designs historically have tended to use experience, intuition, and subjective judgment in siting monitoring stations only sporadically. Better design procedures for optimizing monitoring systems with respect to multiple criteria decision analysis had rarely been put into practice up front when the needs for intensive monitoring became critical. This paper describes a systematic relocation strategy that is organized to identify several significant planning objectives and consider a series of inherent constraints simultaneously. The planning objectives considered in this analysis are designed to enhance the detection possibility for lower compliance areas, reflect the emphasis for different attainable water uses at different locations, promote the potential detection for the lower degradation areas of pollutants, increase the protection degree of those areas with higher population density in the proximity of the river system, and strengthen the pre‐warning capability of water quality for water intakes. The constraint set contains the limitations of budget, the equity implication, and the detection sensitivity in the water environment. A case study in the Kao‐Ping River Basin, South Taiwan, demonstrates the application potential of this methodology based on a seamless integration between the optimization and the simulation models. It enables identification of the optimal locational pattern stepwise using the embedded screening and sequencing capacity in a compromise programming model. However, a well calibrated and verified water quality model is an indispensable tool in support of this multiobjective evaluation. Extra sampling procedures become necessary for the sites with sparse environmental information. Comparison of planning outcomes of compromise programming is made against previously achieved analyses by using weighted programming and fuzzy programming.     

Multivariate statistical characterization of water quality in Lake Lanier, Georgia, USA

Watershed monitoring programs depend on water quality characterization data collected for many parameters, at many times and places, and with limited resources. Our objective is to present a strategy that reduces the measured parameters, locations, and frequency without compromising the quality of the monitoring program. One year of twice-monthly (growing season) and monthly (dormant season) water quality data collected from 17 lake and 10 tributary sites are used in conjunction with multivariate statistical techniques to improve the utility of collected data by identifying key parameters and monitoring locations. Factor analysis shows that tributary water quality data consists of three components-stormwater runoff, municipal and industrial discharges, and ground water-which can be distinguished using total suspended solids, total dissolved solids, and alkalinity plus soluble reactive P, respectively. Lake water quality characterization is more ambiguous than tributary water quality characterization, but factor analysis indicates that anoxia associated with lake stratification is the largest source of lake water quality variation, followed by nutrient abundance, and finally by biomass abundance. Cluster analysis suggests that tributary and lake monitoring stations can be consolidated. Reducing the number of parameters and stations frees up resources for increased monitoring elsewhere.     

A Multi‐Scale Analysis of Low‐Rise Apartment Water Demand through Integration of Water Consumption,Land Use,and Demographic Data

Over the past decades, multi‐unit housing developments have been vastly expanded across urban areas due to the population growth. To properly supply water to this growing sector, it is essential to understand the determinants of its water use. However, this task has largely remained unexplored through the empirical study of water demand mainly due to the scarcity of data in this sector. This study integrated apartment water consumption, property characteristics, weather, water pricing, and census microdata to overcome this issue. Using a rich source of GIS‐based urban databases in Auckland, New Zealand, the study developed a large dataset containing the information of 18,000 low‐rise apartments to evaluate the determinants of water use both in the household scale and aggregated scale. The household‐scale demand analysis helped to assess the heterogeneity in responses to the demand drivers specifically water price across different consumer groups, whereas the aggregated analysis revealed the determinants behind the spatial variation in water demand at the census area unit level. Through applying panel data models, the study revealed the household size as the most important determinant of apartment water use in Auckland, where other socioeconomic factors, building features, and water pricing were not significant determinants. This knowledge of determinants of water demand can help water planners to better manage water demand in the compact urban environments.     

A Multi‐Scale Soil Moisture Monitoring Strategy for California: Design and Validation

A multi‐scale soil moisture monitoring strategy for California was designed to inform water resource management. The proposed workflow classifies soil moisture response units (SMRUs) using publicly available datasets that represent soil, vegetation, climate, and hydrology variables, which control soil water storage. The SMRUs were classified, using principal component analysis and unsupervised K‐means clustering within a geographic information system, and validated, using summary statistics derived from measured soil moisture time series. Validation stations, located in the Sierra Nevada, include transect of sites that cross the rain‐to‐snow transition and a cluster of sites located at similar elevations in a snow‐dominated watershed. The SMRUs capture unique responses to varying climate conditions characterized by statistical measures of central tendency, dispersion, and extremes. A topographic position index and landform classification is the final step in the workflow to guide the optimal placement of soil moisture sensors at the local‐scale. The proposed workflow is highly flexible and can be implemented over a range of spatial scales and input datasets can be customized. Our approach captures a range of soil moisture responses to climate across California and can be used to design and optimize soil moisture monitoring strategies to support runoff forecasts for water supply management or to assess landscape conditions for forest and rangeland management.     

Multicriteria Relocation Analysis of an Off-site Radioactive Monitoring Network for a Nuclear Power Plant

Due to increasing environmental consciousness in most countries, every utility that owns a commercial nuclear power plant has been required to have both an on-site and off-site emergency response plan since the 1980s. A radiation monitoring network, viewed as part of the emergency response plan, can provide information regarding the radiation dosage emitted from a nuclear power plant in a regular operational period and/or abnormal measurements in an emergency event. Such monitoring information might help field operators and decision-makers to provide accurate responses or make decisions to protect the public health and safety. This study aims to conduct an integrated simulation and optimization analysis looking for the relocation strategy of a long-term regular off-site monitoring network at a nuclear power plant. The planning goal is to downsize the current monitoring network but maintain its monitoring capacity as much as possible. The monitoring sensors considered in this study include the thermoluminescence dosimetry (TLD) and air sampling system (AP) simultaneously. It is designed for detecting the radionuclide accumulative concentration, the frequency of violation, and the possible population affected by a long-term impact in the surrounding area regularly while it can also be used in an accidental release event. With the aid of the calibrated Industrial Source Complex–Plume Rise Model Enhancements (ISC-PRIME) simulation model to track down the possible radionuclide diffusion, dispersion, transport, and transformation process in the atmospheric environment, a multiobjective evaluation process can be applied to achieve the screening of monitoring stations for the nuclear power plant located at Hengchun Peninsula, South Taiwan. To account for multiple objectives, this study calculated preference weights to linearly combine objective functions leading to decision-making with exposure assessment in an optimization context. Final suggestions should be useful for narrowing the set of scenarios that decision-makers need to consider in this relocation process.     

The Watershed Flow and Allocation Model: An NHDPlus‐Based Watershed Modeling Approach for Multiple Scales and Conditions

The Watershed Flow and Allocation model (WaterFALL^®) provides segment‐specific, daily streamflow at both gaged and ungaged locations to generate the hydrologic foundation for a variety of water resources management applications. The model is designed to apply across the spatially explicit and enhanced National Hydrography Dataset (NHDPlus) stream and catchment network. To facilitate modeling at the NHDPlus catchment scale, we use an intermediate‐level rainfall‐runoff model rather than a complex process‐based model. The hydrologic model within WaterFALL simulates rainfall‐runoff processes for each catchment within a watershed and routes streamflow between catchments, while accounting for withdrawals, discharges, and onstream reservoirs within the network. The model is therefore distributed among each NHDPlus catchment within the larger selected watershed. Input parameters including climate, land use, soils, and water withdrawals and discharges are georeferenced to each catchment. The WaterFALL system includes a centralized database and server‐based environment for storing all model code, input parameters, and results in a single instance for all simulations allowing for rapid comparison between multiple scenarios. We demonstrate and validate WaterFALL within North Carolina at a variety of scales using observed streamflows to inform quantitative and qualitative measures, including hydrologic flow metrics relevant to the study of ecological flow management decisions.     

GIS‐Based Predictive Models of Hillslope Runoff Generation Processes1

Abstract: Successful nonpoint source pollution control using best management practice placement is a complex process that requires in‐depth knowledge of the locations of runoff source areas in a watershed. Currently, very few simulation tools are capable of identifying critical runoff source areas on hillslopes and those available are not directly applicable under all runoff conditions. In this paper, a comparison of two geographic information system (GIS)‐based approaches: a topographic index model and a likelihood indicator model is presented, in predicting likely locations of saturation excess and infiltration excess runoff source areas in a hillslope of the Savoy Experimental Watershed located in northwest Arkansas. Based on intensive data collected from a two‐year field study, the spatial distributions of hydrologic variables were processed using GIS software to develop the models. The likelihood indicator model was used to produce probability surfaces that indicated the likelihood of location of both saturation and infiltration excess runoff mechanisms on the hillslope. Overall accuracies of the likelihood indicator model predictions varied between 81 and 87% for the infiltration excess and saturation excess runoff locations respectively. On the basis of accuracy of prediction, the likelihood indicator models were found to be superior (accuracy 81‐87%) to the predications made by the topographic index model (accuracy 69.5%). By combining statistics with GIS, runoff source areas on a hillslope can be identified by incorporating easily determined hydrologic measurements (such as bulk density, porosity, slope, depth to bed rock, depth to water table) and could serve as a watershed management tool for identifying critical runoff source areas in locations where the topographic index or other similar methods do not provide reliable results.     

JUSTIFICATION FOR A REDUCTION IN THE CREST-STAGE GAGE PROGRAM IN LOUISIANA1

The crest-stage gage program in Louisiana was evaluated to determine if the data were adequate for use in developing regional flood-frequency equations and to determine if any crest-stage gage stations could be discontinued. An abundance of data at many crest-stage gage stations and a lack of data for urban areas and flat-slope areas indicated a need for a shift in the number, type, and locations of gages. Correlations and comparisons of annual peak discharges and watershed characteristics of 96 existing stations resulted in the elimination of 72 stations and the addition of one new station, reducing the total network to 25 stations that could be used for future flood-frequency analyses. The adequacy of the reduced network for development and verification of regional flood-frequency equations was tested by comparing a set of regional flood-frequency equations developed using data from the full network with a set developed using data from the reduced network. The results indicate that the crest-stage gage network can be reduced to 25 stations and still provide adequate information for future flood-frequency analyses.     

Application of Tabu Search Algorithm With a Coupled AnnAGNPS‐CCHE1D Model to Optimize Agricultural Land Use1

Abstract: A principal contributor to soil erosion and nonpoint source pollution, agricultural activities have a major influence on the environmental quality of a watershed. Impact of agricultural activities on the quality of water resources can be minimized by implementing suitable agriculture land‐use types. Currently, land uses are designed (location, type, and operational schedule) based on field study results, and do not involve a science‐based approach to ensure their efficiency under particular regional, climatic, geological, and economical conditions. At present, there is a real need for new methodologies that can optimize the selection, design, and operation of agricultural land uses at the watershed scale by taking into account environmental, technical, and economical considerations, based on realistic simulations of watershed response. In this respect, the present study proposes a new approach, which integrates computational modeling of watershed processes, fluvial processes in the drainage network, and modern heuristic optimization techniques to design cost effective land‐use plans. The watershed model AnnAGNPS and the channel network model CCHE1D are linked together to simulate the sediment and pollutant transport processes. Based on the computational results, a multi‐objective function is set up to minimize soil losses, nutrient yields, and total associated costs, while the production profits from agriculture are maximized. The selected iterative optimization algorithm uses adaptive Tabu Search heuristic to flip (switching from one alternative to another) land‐change variables. USDA’s Goodwin Creek experimental watershed, located in Northern Mississippi, is used to demonstrate the capabilities of the proposed approach. The results show that the optimized land‐use design with BMPs using an integrated approach at the watershed level can provide efficient and cost‐effective conservation of the environmental quality by taking into account both productivity and profitability.     

Automatic Calibration of Hydrologic Models With Multi‐Objective Evolutionary Algorithm and Pareto Optimization1

Abstract: In optimization problems with at least two conflicting objectives, a set of solutions rather than a unique one exists because of the trade‐offs between these objectives. A Pareto optimal solution set is achieved when a solution cannot be improved upon without degrading at least one of its objective criteria. This study investigated the application of multi‐objective evolutionary algorithm (MOEA) and Pareto ordering optimization in the automatic calibration of the Soil and Water Assessment Tool (SWAT), a process‐based, semi‐distributed, and continuous hydrologic model. The nondominated sorting genetic algorithm II (NSGA‐II), a fast and recent MOEA, and SWAT were called in FORTRAN from a parallel genetic algorithm library (PGAPACK) to determine the Pareto optimal set. A total of 139 parameter values were simultaneously and explicitly optimized in the calibration. The calibrated SWAT model simulated well the daily streamflow of the Calapooia watershed for a 3‐year period. The daily Nash‐Sutcliffe coefficients were 0.86 at calibration and 0.81 at validation. Automatic multi‐objective calibration of a complex watershed model was successfully implemented using Pareto ordering and MOEA. Future studies include simultaneous automatic calibration of water quality and quantity parameters and the application of Pareto optimization in decision and policy‐making problems related to conflicting objectives of economics and environmental quality.     

Approximating SWAT Model Using Artificial Neural Network and Support Vector Machine1

Abstract: With the popularity of complex, physically based hydrologic models, the time consumed for running these models is increasing substantially. Using surrogate models to approximate the computationally intensive models is a promising method to save huge amounts of time for parameter estimation. In this study, two learning machines [Artificial Neural Network (ANN) and support vector machine (SVM)] were evaluated and compared for approximating the Soil and Water Assessment Tool (SWAT) model. These two learning machines were tested in two watersheds (Little River Experimental Watershed in Georgia and Mahatango Creek Experimental Watershed in Pennsylvania). The results show that SVM in general exhibited better generalization ability than ANN. In order to effectively and efficiently apply SVM to approximate SWAT, the effect of cross‐validation schemes, parameter dimensions, and training sample sizes on the performance of SVM was evaluated and discussed. It is suggested that 3‐fold cross‐validation is adequate for training the SVM model, and reducing the parameter dimension through determining the parameter values from field data and the sensitivity analysis is an effective means of improving the performance of SVM. As far as the training sample size, it is difficult to determine the appropriate number of samples for training SVM based on the test results obtained in this study. Simple examples were used to illustrate the potential applicability of combining the SVM model with uncertainty analysis algorithm to save efforts for parameter uncertainty of SWAT. In the future, evaluating the applicability of SVM for approximating SWAT in other watersheds and combining SVM with different parameter uncertainty analysis algorithms and evolutionary optimization algorithms deserve further research.     

AN OBJECTWE TEST FOR HYDROLOGIC SCALE1

ABSTRACT: Improving the reliability of parametric hydrologic models (sometimes called cenceptual rainfall-runoff models) in the continuous simulation of runoff from ungaged catchments has been frustrated by difficulties in estimating model parameters from catchment characteristics. An underlying problem is that these models use parameters to represent catchments as a whole, whereas data on catchment characteristics are collected at multiple field locations and are difficult to transform into one measure of collective impact. Subdividing the catchment and calibrating a stochastic parametric model to estimate distributions for the parameters that covered the range of observed streamflow values was found to improve the simulations. This paper presents an optimization of the amount of subdivision to use in simulation with a version of the Stanford Watershed Model using available climatological data. The calibration process assumes that catchment heterogeneity introduces errors that can be reduced by calibrating parameters as spatial distributions rather than single values. Calibrations for three diverse small gaged catchments located in California and in Virginia found the optimal number of subdivisions to range from 4 to 25 and the optimal scale to range from 0.3 to 2.1 mi².     

AN OPTIMIZATION TECHNIQUE ADAPTED TO USDAHL-74 REVISED MODEL OF WATERSHED HYDROLOGY1

ABSTRACT. A direct search optimization routine has been added to USDAHL-74 Revised Model of Watershed Hydrology. It provides an objective method to determine values of selected parameters to give the optimum match between observed and computed hydrologic data. Examination of intermediate results also gives information on the sensitivity of the model to changes in the parameter being optimized.