Search

This study assessed the potential impact of climate change on spatiotemporal distribution of blue and green water resources in the Beijiang River basin, southern China, using the SWAT hydrological model and CORDEX-EAS regional climate models (RCMs). The outputs of three RCMs (namely RMOH, RMPI and RNCC) under two emission scenarios (RCP 2.6 and RCP 8.5) were bias-corrected by using the quantile delta mapping method for the control (1975–2004), near future (2021–2050) and far future (2061–2090) periods. Driven by the corrected climate variables, future blue and green water were assessed by using the SWAT model calibrated with streamflow data. The results indicated that the green water flow in the northwest basin predicted by RMOH under RCP 2.6 would increase up to about 2.7% in 2061–2090; RNCC suggests that most of the basin would experience the most significant reduction in green water storage (over 30%) under RCP 8.5 in the far future. For blue water, the ensemble mean of three RCMs indicates a slight decreasing trend in the northern part of the basin and an increasing one in the southern part under RCP 8.5 in the far future. Our findings could help watershed managers evaluate future hydrological risks and design appropriate adaptation strategies. Ministry of Education (MOE) Published version This work was supported by the National Natural Science Foundation of China (52009022) and AcRF Tier 1 project (2019- T1-001-160) from the Ministry of Education (MOE), Singapore.

This study investigated the temporal rainfall pattern in order to facilitate rainfall design, which normally requires a good understanding of the temporal patterns of rainstorm events. The analysis employed a storm-event-based approach using the concept of inter-event time definition (IETD) and rainfall depth/duration/intensity thresholds. The 5-min rainfall data at three raingauge stations were analysed to determine representative quartiles of a design storm in a tropical city. The temporal characteristics of the design storm could be determined from the rainfall depth ratios of consecutive peak rainfalls for each interval of storm duration, and time to the first peak rainfall depending on each quartile’s rainstorm events. The determination of the quantile distribution of tropical rainfall could help improve the representativeness of design rainfall and facilitate rainfall–runoff modelling for urban flood control in a tropical region. Ministry of Education (MOE) This work was partly supported by Academic Research Fund Tier 1 from the Ministry of Education (MOE), Singapore (2019-T1-001-160), and partly by the Chung-Ang University Research Grants in 2019.

Urban areas are becoming increasingly vulnerable to extreme storms and flash floods, which could be more damaging under climate change. This study presented an integrated framework for assessing climate change impact on extreme rainfall and urban drainage systems by incorporating a number of statistical and modelling techniques. Starting from synthetic future climate data generated by the stochastic weather generator, the simple scaling method and the Huff rainfall design were adopted for rainfall disaggregation and rainfall design. After having obtained 3-min level designed rainfall information, the urban hydrological model (i.e., Storm Water Management Model) was used to carry out the runoff analysis. A case study in a tropical city was used to demonstrate the proposed framework. Particularly, the impact of selecting different general circulation models and Huff distributions on future 1-h extreme rainfall and the performance of the urban drainage system were investigated. It was revealed that the proposed framework is flexible and easy to implement in generating temporally high-resolution rainfall data under climate model projections and offers a parsimonious way of assessing urban flood risks considering the uncertainty arising from climate change model projections, downscaling and rainfall design. Ministry of Education (MOE) Nanyang Technological University Published version This project was supported by Start-Up Grant (M4081327.030) from the School of Civil and Environmental Engineering, Nanyang Technological University, Singapore and in part by the AcRF Tier 1 project (M4012228.030) from the Ministry of Education (MOE), Singapore.

To improve the convergence of multiple-site weather generators (SWGs) based on the brute force algorithm (MBFA), a genetic algorithm (GA) is proposed to search the overall optimal correlati...

Seeking cost-effective design of urban hydrological facilities and drainage systems is an important task for many city planners. However, such a process has always been complicated with intrinsic uncertainties. This work presented an integrated fuzzy simulation-optimization model (FSOM) for supporting Low Impact Development (LID) design under model uncertainties. Various LID implementation schemes involving green roof, bio-retention cell, and permeable pavement were simulated through an urban hydrological model. Three model parameters were assumed as fuzzy sets. In a case study, fuzzy simulation (FS) and genetic algorithm (GA) were employed to search the optimal schemes of LIDs under various confidence levels of satisfying flood control constraints. Comparison of FSOM to traditional deterministic and stochastic models were also carried out. It was shown that FSOM could offer a flexible way of defining and assessing uncertainties associated with hydrological modeling and generate solutions that were comparable to those from either deterministic or stochastic models. However, FSOM also showed limitation of high computational requirement.

A probabilistic flood risk modelling chain is proposed for flood risk analysis with consideration of spatial spreading and temporal clustering of the flood events. The proposed method consists of (1) a continuous simulation of long-term climatic and hydrologic fields, (2) Monte-Carlo simulations of a 1D river and inundation model and (3) a probabilistic loss model. In the first stage, three multisite multivariate weather generators were tested and a K-nearest neighbor weather generator (KNN-CAD v4) was found most suitable to reproduce hydrological extremes. The methodology was then applied to identify variations in risk between spatially coherent and simplified event sets generated from 5000 years of synthetic data across four river basins in Uttarakhand, India. The results showed that preserving the spatial coherency of regional events lead to a spreading of the T-year local event distributions on both sides of a uniform regional T-year return period. The specificity of the exposure and the correlations between local events translated into minimal (4%) differences in losses at high return periods (> 50 year) but larger difference (20%) at low return periods (5 to 10-year). The redistribution of loss frequencies led to negligible differences in the Annual Average Losses between the different event sets. In this application, a simplified event set structure with uniform but appropriate T-year return periods proved to capture reasonably well the averaged risk metrics regionally. This study illustrates the complexity of estimating regional flood risk accumulation and exemplifies how a probabilistic risk model chain with continuous simulation can provide a more detailed picture of flood risks. Economic Development Board (EDB) Nanyang Technological University The study was funded by the Economic Development Board (EDB) of Singapore (IPP Scholarship) and the DHI-NTU Research Centre and Education Hub. The work is closely affiliated with the Disaster Risk Assessment of Uttarakhand, an initiative of the Uttarakhand Disaster Recovery Programme (UDRP) funded by the World Bank.

The potential global warming impact on the extreme and mean streamflow of the Upper Blue Nile (UBN) River Basin was projected for the 2050s and 2080s by three hydrological models [Nedbor-Af...

This study investigated the frequency analysis of precipitation indices of maximum daily precipitation amount (Rx1day), maximum 5-day cumulative precipitation amount (Rx5day), and maximum length of consecutive dry days (CDD) using various bias-correction schemes over mainland Southeast Asia (MSEA). The bias-correction schemes were based on quantile delta mapping (QDM) with different strategies in selecting the data series, including monthly correction (MC), annual correction (AC), peak correction using annual raw data (PCRAW), and annual correction using extreme-index data series (ACEX). Two regional climate model (RCM) outputs from CORDEX-SEA and APHRODITE rainfall datasets were used as modelled and observed data for methodology verification. The study first compared four frequency analysis methods based on generalized extreme value (GEV), Gumbel (GB), Log-Pearson Type III (LP3), and Lognormal (LOGN) distributions, at 9 selected sites over MSEA, and found that LP3 was the best choice for frequency analysis for the study region. Then, the study compared the four bias-correction schemes at both individual sites and all grids over the entire MSEA and indicated that the PCRAW was the best performer in terms of bias-correction for frequency information. Finally, the study gave an ensembled projection of future extreme indices with 200-yr return based on seven CORDEX-SEA RCMs and suggested a general increasing trend of all indices over MSEA. The study explored the effect of uncertainty originated from adopting various bias-correction schemes in mapping future frequency of precipitation indices and is valuable in revealing the spatiotemporal distribution of precipitation extremes over large areas under climate change which is important for flood/drought risk assessment and adaptation planning. Ministry of Education (MOE) This study was supported by AcRF Tier 1 project (2019-T1-001-160) from Ministry of Education (MOE), Singapore. We acknowledge the data support from APHRODITE (http://aphrodite.st.hirosaki-u.ac.jp/) and the Southeast Asia Regional Climate Downscaling (SEACLID)/Coordinated Regional Climate Downscaling Experiment for Southeast Asia(CORDEX-SEA) project (which was funded by ARCP2014-07CMY-Tangang and ARCP2015-04CMY-Tangang).

A Gini-coefficient based stochastic optimization (GBSO) model was developed by integrating the hydrological model, water balance model, Gini coefficient and chance-constrained programming (CCP) into a general multi-objective optimization modeling framework for supporting water resources allocation at a watershed scale. The framework was advantageous in reflecting the conflicting equity and benefit objectives for water allocation, maintaining the water balance of watershed, and dealing with system uncertainties. GBSO was solved by the non-dominated sorting Genetic Algorithms-II (NSGA-II), after the parameter uncertainties of the hydrological model have been quantified into the probability distribution of runoff as the inputs of CCP model, and the chance constraints were converted to the corresponding deterministic versions. The proposed model was applied to identify the Pareto optimal water allocation schemes in the Lake Dianchi watershed, China. The optimal Pareto-front results reflected the tradeoff between system benefit ( α SB ) and Gini coefficient ( α G ) under different significance levels (i.e. q ) and different drought scenarios, which reveals the conflicting nature of equity and efficiency in water allocation problems. A lower q generally implies a lower risk of violating the system constraints and a worse drought intensity scenario corresponds to less available water resources, both of which would lead to a decreased system benefit and a less equitable water allocation scheme. Thus, the proposed modeling framework could help obtain the Pareto optimal schemes under complexity and ensure that the proposed water allocation solutions are effective for coping with drought conditions, with a proper tradeoff between system benefit and water allocation equity.

A comprehensive flood risk assessment should aim not only at quantifying uncertainties but also the variability of risk over time. In this study, an efficient modelling framework was proposed to perform probabilistic hazard and risk analysis in dike-protected river systems accounting for morphological variability and uncertainty. The modelling framework combined the use of: (1) continuous synthetic discharge forcing, (2) a stochastic dike breach model dynamically coupled to a stochastic unsteady one-dimensional hydraulic model (MIKE1D) describing river flows, (3) a catalogue of pre-run probabilistic inundation maps (MIKE SHE) and (4) a damage and loss model (CAPRA). The methodology was applied using continuous simulations to a 45-km reach of the Upper Koshi River, Nepal, to investigate the changes in breach and flood hazards and subsequent risks after 2 and 5 years of probable river bed aggradation. The study results indicated an increase in annual average loss of 4% per year driven by changes in loss distribution in the most frequent loss return periods (20–500 years). The use of continuous simulations and dike breach model also provided a more robust estimation of risk metrics as compared to traditional binary treatment of flood defence and/or the direct association of flow with loss return periods. The results were helpful to illustrate the potential impacts of dynamic river morphology, dike failure and continuous simulation and their significance when devising flood risk study methodologies.

In this study, a storm event-based frequency analysis method was proposed to mitigate the limitations of conventional rainfall depth–duration–frequency (DDF) analysis. The proposed method takes the number, rainfall depth, and duration of rainstorm events into consideration and is advantageous in estimation of more realistic rainfall quantiles for a given return period. For the purpose of hydraulics design, the rainfall depth thresholds are incorporated to retrieve the rainstorm events for estimating design rainfalls. The proposed method was tested against the observed rainfall data from 1961 to 2010 at Seoul, Korea and the computed rainfall quantiles were compared with those estimated using the conventional frequency analysis method. The study results indicated that the conventional method was likely to overestimate the rainfall quantiles for short rainfall durations. It represented that the conventional method could reflect rainfall characteristics of actual rainstorm events if longer durations (like 24 hours) were considered for estimation of design rainfalls.

Current state-of-the-art online neuro fuzzy models (NFMs) such as DENFIS (Dynamic Evolving Neural-Fuzzy Inference System) have been used for runoff forecasting. Online NFMs adopt a local learning approach and are able to adapt to changes continuously. The DENFIS model however requires upper/lower bound for normalization and also the number of rules increases monotonically. This requirement makes the model unsuitable for use in basins with limited data, since a priori data is required. In order to address this and other drawbacks of current online models, the Generic Self-Evolving Takagi-Sugeno-Kang (GSETSK) is adopted in this study for forecast applications in basins with limited data. GSETSK is a fully-online NFM which updates its structure and parameters based on the most recent data. The model does not require the need for historical data and adopts clustering and rule pruning techniques to generate a compact and up-to-date rule-base. GSETSK was used in two forecast applications, rainfall-runoff (a catchment in Sweden) and river routing (Lower Mekong River) forecasts. Each of these two applications was studied under two scenarios: (i) there is no prior data, and (ii) only limited data is available (1 year for the Swedish catchment and 1 season for the Mekong River). For the Swedish Basin, GSETSK model results were compared to available results from a calibrated HBV (Hydrologiska Byrans Vattenbalansavdelning) model. For the Mekong River, GSETSK results were compared against the URBS (Unified River Basin Simulator) model. Both comparisons showed that results from GSETSK are comparable with the physically based models, which were calibrated with historical data. Thus, even though GSETSK was trained with a very limited dataset in comparison with HBV or URBS, similar results were achieved. Similarly, further comparisons between GSETSK with DENFIS and the RBF (Radial Basis Function) models highlighted further advantages of GSETSK as having a rule-base (compared to opaque RBF) which is more compact, up-to-date and more easily interpretable.

Optimization techniques have emerged as robust tools to aid the planning and design of urban drainage facilities in cost-effective ways. Such an effort was traditionally aided by heuristic methods (like genetic algorithm), which was generally time-consuming and also challenging in reaching convergence for large-scale problems with wide decision spaces. This study proposed a novel optimization method, denoted as two-level optimization (TO) scheme, for supporting rainwater storage pond design in an urban drainage system. Polynomial regression models were established as surrogate models to facilitate the solution of the optimization framework using traditional iteration algorithm. The TO scheme firstly sought the optimal layout of storage ponds on tributary sub-watersheds, and then proceeded to that of the mainstream one to yield the final solution. Through a case study, the TO scheme was compared with the traditional global optimization (GO) scheme where the physical simulation model was dynamically linked with genetic algorithm (GA) to seek the global optimal solution. The performance of two schemes under different constraint settings was analyzed. Effects of related issues such as start-point selection and mainstream design on tributary sub-watersheds were also discussed. The results showed that the proposed TO scheme is a prominent alternative to the traditional GO scheme to support urban water managers for a more science-based decision making towards storage pond implementation in large-scale practical problems. Ministry of Education (MOE) The research work was supported by Singapore’s Ministry of Education (MOE) AcRF Tier 1 Project (Ref No. RG170/16; WBS No.: 4011766.030).

This review article discusses the climate, water resources and historical droughts of Africa, drought indices, vulnerability, impact of global warming and land use for drought-prone regions in West, southern and the Greater Horn of Africa, which have suffered recurrent severe droughts in the past. Recent studies detected warming and drying trends in Africa since the mid 20th century. Based on the Fourth Assessment Report of the Intergovernmental Panel on Climate Change and the Coupled Model Intercomparison Project Phase 5 (CMIP5), both northern and southern Africa are projected to experience drying, such as decreasing precipitation, runoff and soil moisture in the 21st century and could become more vulnerable to the impact of droughts. The daily maximum temperature is projected to increase by up to 8°C (RCP8.5 of CMIP5), precipitation indices such as total wet day precipitation (PRCPTOT) and heavy precipitation days (R10 mm) could decrease, while warm spell duration (WSDI) and consecutive dry days...

Summary An integrated statistical and data-driven (ISD) framework was proposed for analyzing river flows and flood frequencies in the Duhe River Basin, China, under climate change. The proposed framework involved four major components: (i) a hybrid model based on ASD (Automated regression-based Statistical Downscaling tool) and KNN (K-nearest neighbor) was used for downscaling rainfall and CDEN (Conditional Density Estimate Network) was applied for downscaling minimum temperature and relative humidity from global circulation models (GCMs) to local weather stations; (ii) Bayesian neural network (BNN) was used for simulating monthly river flows based on projected weather information; (iii) KNN was applied for converting monthly flow to daily time series; (iv) Generalized Extreme Value (GEV) distribution was adopted for flood frequency analysis. In this study, the variables from CGCM3 A2 and HadCM3 A2 scenarios were employed as the large-scale predictors. The results indicated that the maximum monthly and annual runoffs would both increase under CGCM3 and HadCM3 A2 emission scenarios at the middle and end of this century. The flood risk in the study area would generally increase with a widening uncertainty range. Compared with traditional approaches, the proposed framework takes the full advantages of a series of statistical and data-driven methods and offers a parsimonious way of projecting flood risks under climatic change conditions.

This study utilized a long-term (1951–2014) moderate-resolution observed gridded (0.5°×0.5°) rainfall dataset to analyze long-term and short-term rainfall variability and the seasonality in Southeast Asia (SEA). The previous studies for this region revealed significant enhancement in extreme events and seasonal variabilities in rainfall. In this study, the frequency and intensity-based rainfall extremity analysis was carried out by utilizing a number of rainfall extreme indices (e.g. Dry Days, Wet Days, 90P, 95P, and 99P). A seasonal non-parametric Mann-Kendall test was applied to detect significant increase and decrease of rainfall. The possible causes for the acceleration and variabilities of rainfall were investigated by analyzing the impact of large-scale ocean-atmospheric climate interactions through global climate indices such as El Nino (3.4 and 4.0), Southern Oscillation Index (SOI), Madden-Julian Oscillation (MJO), global mean land and ocean temperature index, QBO and NOI. Our results displayed a significant increase of rainfall amount in most of SEA regions, while a minimal decrease was also found over some regions. The rainfall indices displayed substantial intra-decadal variabilities over the SEA region. The intra-decadal percent of change analysis based on rainfall extreme indices revealed significant changes in rainfall extremes over SEA, which also evidenced the computed acceleration in rainfall over the region. The correlation analysis based findings exhibited that a strong correlation existed between the SEA rainfall and the large-scale global climate indices generated based on the ocean-atmosphere climate interactions. After regionalization of the SEA, the global climate indices based results described the spatiotemporal variabilities in rainfall, which were computed over SEA regions. The global climate indices such as El Nino (3.4 & 4.0), SOI, NOI and MJO suggested significant correlations with the seasonal rainfall and the rainfall extreme indices and highlighted that their collinearities vary within regions as per variations in the monsoon.

A generalized likelihood uncertainty estimation (GLUE) framework coupling with artificial neural network (ANN) models in two surrogate schemes (i.e. GAE-S1 and GAE-S2) was proposed to improve the efficiency of uncertainty assessment in flood inundation modelling. The GAE-S1 scheme was to construct an ANN to approximate the relationship between model likelihoods and uncertain parameters for facilitating sample acceptance/rejection instead of running the numerical model directly; thus, it could speed up the Monte Carlo simulation in stochastic sampling. The GAE-S2 scheme was to establish independent ANN models for water depth predictions to emulate the numerical models; it could facilitate efficient uncertainty analysis without additional model runs for locations concerned under various scenarios. The results from a study case showed that both GAE-S1 and GAE-S2 had comparable performances to GLUE in terms of estimation of posterior parameters, prediction intervals of water depth, and probabilistic i...

Fuzzy programming model has been widely used in water resources management, but its applicability has been significantly restricted in dealing with triangular or trapezoidal shaped fuzzy sets due to intrinsic complexity in converting fuzzy constraints into their deterministic forms. In this study, a novel superiority-inferiority-based sequential fuzzy programming (SISFP) model was proposed for supporting water supply–demand analysis under uncertainty. The SISFP method could transform fuzzy objective function and constraints with general-shaped fuzzy coefficients into their crisp equivalent by using fuzzy superiority and inferiority measures. The water supply–demand management system in Tianjin Binhai New Area, China, consisting of five sources of water, five water users at three districts (i.e. Tanggu, Hanggu, and Dagang), was used for methodology demonstration. The proposed model could effectively address the complex nature of fuzzy characterization of water-transfer safety factor, wastewater reclamation rate, the net benefits derived from water, and water-saving rate of the system; and also take demand management measures into consideration. The obtained solutions have sought a well balance among the water availability, water demand, adoption of water-saving measures, and benefit/cost of each water users. The advantage and necessity of SISFP in dealing with general-shaped fuzzy parameters were further verified by comparing to fuzzy models with both deterministic and specially-shaped fuzzy parameters.

A stochastic optimization model for urban drainage design (SODD) was proposed in this study to help analyze the tradeoff between investment and acceptable flood damage in urban drainage designs considering effects of both uncertainty and climate change. The simulation model (i.e. SWMM), driven by designed rainfall either from existing Intensity-Duration-Frequency (IDF) curves or future ones subjected to climate change conditions, was used to simulate flooding scenarios. The generalized uncertainty analysis estimation (GLUE) and Monte Carlo simulation methods were employed to quantify the system reliability which was adopted in the constraints of the optimization model. The results from a case study showed that the deterministic optimization was computationally efficient with no randomness encountered in hydrological simulation, although its solution was hardly reliable in achieving the target for flood mitigation. The stochastic version, on the other hand, could offer richer information on system reliability and help managers make a more robust decision. The results also revealed that the rainfall extremes under the impact of climate change could significantly affect system investment. The proposed method is advantageous in facilitating cost-effective planning towards a risk-based drainage design in light of various complexities.

An uncertainty assessment framework based on Karhunen–Loevè expansion (KLE) and probabilistic collocation method (PCM) was introduced to deal with flood inundation modelling under uncertainty. The Manning's roughness for channel and floodplain were treated as 1D and 2D, respectively, and decomposed by KLE. The maximum flow depths were decomposed by the 2nd-order PCM. Through a flood modelling case with steady inflow hydrographs based on five designed testing scenarios, the applicability of KLE-PCM was demonstrated. The study results showed that the Manning's roughness assumed as a 1D/2D random field could efficiently alleviate the burden of random dimensionality within the analysis framework, and the introduced method could significantly reduce repetitive runs of the physical model as required in the traditional Monte Carlo simulation (MCS). The study sheds some light on reducing the computational burden associated with flood modelling under uncertainty which is useful for the related damage quantification and risk management.

This paper presents the application of an improved particle swarm optimization (PSO) technique for training an artificial neural network (ANN) to predict water levels for the Heshui Watershed, China. Daily values of rainfall and water levels from 1988 to 2000 were first analyzed using ANNs trained with the conjugate gradient, gradient descent, and Levenberg-Marquardt neural network (LM-NN) algorithms. The best results were obtained from the LM-NN, and these results were then compared with those from PSO-based ANNs, including the conventional PSO neural network (CPSONN) and the improved PSO neural network (IPSONN) with passive congregation. The IPSONN algorithm improves PSO convergence by using the selfish herd concept in swarm behavior. The results show that the PSO-based ANNs performed better than the LM-NN. For models run using a single parameter (rainfall) as input, the root mean square error (RMSE) of the testing data set for IPSONN was the lowest (0.152 m) compared to those for CPSONN (0.161 ...

A generalized likelihood uncertainty estimation (GLUE) method incorporating moving least squares (MLS) with entropy for stochastic sampling (denoted as GLUE-MLS-E) was proposed for uncertainty analysis of flood inundation modelling. The MLS with entropy (MLS-E) was established according to the pairs of parameters/likelihoods generated from a limited number of direct model executions. It was then applied to approximate the model evaluation to facilitate the target sample acceptance of GLUE during the Monte-Carlo-based stochastic simulation process. The results from a case study showed that the proposed GLUE-MLS-E method had a comparable performance as GLUE in terms of posterior parameter estimation and predicted confidence intervals; however, it could significantly reduce the computational cost. A comparison to other surrogate models, including MLS, quadratic response surface and artificial neural networks (ANN), revealed that the MLS-E outperformed others in light of both the predicted confidence interval and the most likely value of water depths. ANN was shown to be a viable alternative, which performed slightly poorer than MLS-E. The proposed surrogate method in stochastic sampling is of practical significance in computationally expensive problems like flood risk analysis, real-time forecasting, and simulation-based engineering design, and has a general applicability in many other numerical simulation fields that requires extensive efforts in uncertainty assessment. Copyright © 2014 John Wiley & Sons, Ltd.

Water-distribution operation is an essential part of an urban water-supply system to deliver high-quality water to consumers. Management of such an operation may involve deliberations of operation cost, system capacity, and environmental restriction, and is convoluted with many forms of uncertainties. In this paper, an integrated fuzzy programming and decision analysis (IFPDA) approach was proposed for a multilayer urban water-distribution system management under uncertainty. The system consisted of two water sources, four treatment plants, seven reservoirs, and seven consuming zones. Uncertain information associated with water demands, water-treatment capacities, water-transfer cost, and leakage rate was described by a trapezoidal-shaped fuzzy set and embedded into a fuzzy programming framework. The balance between the satisfaction degree of achieving the system objective and feasibility level of meeting the related constraints was analyzed by fuzzy inference procedures. The results indicate that...

Developing an efficient and accurate hydrologic forecasting model is crucial to managing water resources and flooding issues. In this study, response surface (RS) models including multiple linear regression (MLR), quadratic response surface (QRS), and nonlinear response surface (NRS) were applied to daily runoff (e.g., discharge and water level) prediction. Two catchments, one in southeast China and the other in western Canada, were used to demonstrate the applicability of the proposed models. Their performances were compared with artificial neural network (ANN) models, trained with the learning algorithms of the gradient descent with adaptive learning rate (ANN-GDA) and Levenberg-Marquardt (ANN-LM). The performances of both RS and ANN in relation to the lags used in the input data, the length of the training samples, long-term (monthly and yearly) predictions, and peak value predictions were also analyzed. The results indicate that the QRS and NRS were able to obtain equally good performance in runoff prediction, as compared with ANN-GDA and ANN-LM, but require lower computational efforts. The RS models bring practical benefits in their application to hydrologic forecasting, particularly in the cases of short-term flood forecasting (e.g., hourly) due to fast training capability, and could be considered as an alternative to ANN. MOE (Min. of Education, S’pore) Accepted version

A systematic downscaling-disaggregation study was conducted over Singapore Island, with an aim to generate high spatial and temporal resolution rainfall data under future climate-change conditions. The study consisted of two major components. The first part was to perform an inter-comparison of various alternatives of downscaling and disaggregation methods based on observed data. This included (i) single-site generalized linear model (GLM) plus K-nearest neighbor (KNN) (S-G-K) vs. multisite GLM (M-G) for spatial downscaling, (ii) HYETOS vs. KNN for single-site disaggregation, and (iii) KNN vs. MuDRain (Multivariate Rainfall Disaggregation tool) for multisite disaggregation. The results revealed that, for multisite downscaling, M-G performs better than S-G-K in covering the observed data with a lower RMSE value; for single-site disaggregation, KNN could better keep the basic statistics (i.e. standard deviation, lag-1 autocorrelation and probability of wet hour) than HYETOS; for multisite disaggregation, MuDRain outperformed KNN in fitting interstation correlations. In the second part of the study, an integrated downscaling-disaggregation framework based on M-G, KNN, and MuDRain was used to generate hourly rainfall at multiple sites. The results indicated that the downscaled and disaggregated rainfall data based on multiple ensembles from HadCM3 for the period from 1980 to 2010 could well cover the observed mean rainfall amount and extreme data, and also reasonably keep the spatial correlations both at daily and hourly timescales. The framework was also used to project future rainfall conditions under HadCM3 SRES A2 and B2 scenarios. It was indicated that the annual rainfall amount could reduce up to 5% at the end of this century, but the rainfall of wet season and extreme hourly rainfall could notably increase. Accepted version

A simulation-based fuzzy optimization method (SFOM) was proposed for regional groundwater pumping management in considering uncertainties. SFOM enhanced the traditional groundwater management models by incorporating a response matrix model (RMM) into a fuzzy chance-constrained programming (FCCP) framework. RMM was used to approximate the input–output relationship between pumping actions and subsurface hydrologic responses. Due to its explicit expression, RMM could be easily embedded into an optimization model to help seek cost-effective pumping solutions. A groundwater management case in Pinggu District of Beijing, China, was used to demonstrate the method's applicability. The study results showed that the obtained system cost and pumping rates would vary significantly under different confidence levels of constraints satisfaction. The decision-makers could identify the best groundwater pumping strategy through analyzing the tradeoff between the risk of violating the related water resources conservation target and the economic benefit. Compared with traditional approaches, SFOM was particularly advantageous in linking simulation and optimization models together, and tackling uncertainties using fuzzy chance constraints.

A joint Monte Carlo and fuzzy possibilistic simulation (MC-FPS) approach was proposed for flood risk assessment. Monte Carlo simulation was used to evaluate parameter uncertainties associated with inundation modeling, and fuzzy vertex analysis was applied for promulgating human-induced uncertainty in flood damage estimation. A study case was selected to show how to apply the proposed method. The results indicate that the outputs from MC-FPS would present as fuzzy flood damage estimate and probabilistic-possibilistic damage contour maps. The stochastic uncertainty in the flood inundation model and fuzziness in the depth-damage functions derivation would cause similar levels of influence on the final flood damage estimate. Under the worst scenario (i.e. a combined probabilistic and possibilistic uncertainty), the estimated flood damage could be 2.4 times higher than that computed from conventional deterministic approach; considering only the pure stochastic effect, the flood loss would be 1.4 times higher. It was also indicated that uncertainty in the flood inundation modeling has a major influence on the standard deviation of the simulated damage, and that in the damage-depth function has more notable impact on the mean of the fitted distributions. Through applying MC-FPS, rich information could be derived under various α-cut levels and cumulative probabilities, and it forms an important basis for supporting rational decision making for flood risk management under complex uncertainties.

Incorporation of uncertainties within an urban water supply management system has been a challenging topic for many years. In this study, an acceptability-index-based two-step interval programming (AITIP) model was developed for supporting urban water supply analysis under uncertainty. AITIP improved upon the traditional two-step interval programming (TIP) through incorporating the acceptability level of constraints violation into the optimization framework. A four-layer urban water supply system, including water sources, treatment facilities, reservoirs, and consuming zones, was used to demonstrate the applicability of proposed method. The results indicated that an AITIP model was valuable to help understand the effects of uncertainties related to cost, constraints and decision maker’s judgment in the water supply network, and capable of assisting urban water managers gain an in-depth insight into the tradeoffs between system cost and constraints-violation risk. Compared with TIP, the solutions from AITIP were of lower degree of uncertainty, making it more reliable to identify effective water supply patterns by adjusting decision variable values within their solution intervals. The study is useful in helping urban water managers to identify cost-effective management schemes in light of uncertainties in hydrology, environment, and decisions. The proposed optimization approach is expected to be applicable for a wide variety of water resources management problems.

A conditional value-at-risk (CVaR) based inexact two-stage stochastic programming (CITSP) model was developed in this study for supporting water resources allocation problems under uncertainty. A CITSP model was formulated through incorporating a CVaR constraint into an inexact two-stage stochastic programming (ITSP) framework, and could be used to deal with uncertainties expressed as not only probability distributions but also discrete intervals. The measure of risks about the second-stage penalty cost was incorporated into the model, such that the trade-off between system economy and extreme expected loss could be analyzed. The developed model was applied to a water resources allocation problem involving a reservoir and three competing water users. The results indicated that the CITSP model performed better than the ITSP model in its capability of reflecting the economic loss from extreme events. Also, it could generate interval solutions within which the decision alternatives could be selected from a flexible decision space. Overall, the CITSP model was useful for reflecting the decision maker’s attitude toward risk aversion and could help seek cost-effective water resources management strategies under complex uncertainties.

A fuzzy parameterized probabilistic analysis (FPPA) method was developed in this study to assess risks associated with environmental pollution-control problems. FPPA integrated environmental transport modeling, fuzzy transformation, probabilistic risk assessment, fuzzy risk quantification into a general risk assessment framework, and was capable of handling uncertainties expressed as fuzzy-parameterized stochastic distributions. The proposed method was applied to two environmental pollution problems, with one being about the point-source pollution in a river system with uncertain water quality parameters and the other being concerned with groundwater contaminant plume from waste landfill site with poorly known contaminant physical properties. The study results indicated that the complex uncertain features had significant impacts on modeling and risk-assessment outputs; the degree of impacts of modeling parameters were highly dependent on the level of imprecision of these parameters. The results also implied that FPPA was capable of addressing vagueness or imprecision associated with probabilistic risk evaluation, and help generate risk outputs that could be elucidated under different possibilistic levels. The proposed method could be used by environmental managers to evaluate trade-offs involving risks and costs, as well as identify management solutions that sufficiently hedge against dual uncertainties.

An inexact double-sided fuzzy chance-constrained programming (IDFCCP) method was developed in this study and applied to an agricultural effluent control management problem. IDFCCP was formulated through incorporating interval linear programming (ILP) into a double-sided fuzzy chance-constrained programming (DFCCP) framework, and could be used to deal with uncertainties expressed as not only possibility distributions associated with both left- and right-hand-side components of constraints but also discrete intervals in the objective function. The study results indicated that IDFCCP allowed violation of system constraints at specified confidence levels, where each confidence level consisted of two reliability scenarios. This could lead to model solutions with high system benefits under acceptable risk magnitudes. Furthermore, the introduction of ILP allowed uncertain information presented as discrete intervals to be communicated into the optimization process, such that a variety of decision alternatives can be generated by adjusting the decision-variable values within their intervals. The proposed model could help decision makers establish various production patterns with cost-effective water quality management schemes under complex uncertainties, and gain in-depth insights into the trade-offs between system economy and reliability.

An integrated simulation-assessment approach (ISAA) was developed in this study to systematically tackle multiple uncertainties associated with hydrocarbon contaminant transport in subsurface and assessment of carcinogenic health risk. The fuzzy vertex analysis technique and the Latin hypercube sampling (LHS) based stochastic simulation approach were combined into a fuzzy-Latin hypercube sampling (FLHS) simulation model and was used for predicting contaminant transport in subsurface under coupled fuzzy and stochastic uncertainties. The fuzzy-rule-based risk assessment (FRRA) was used for interpreting the general risk level through fuzzy inference to deal with the possibilistic uncertainties associated with both FLHS simulations and health-risk criteria. A study case involving health risk assessment for a benzene-contaminated site was examined. The study results demonstrated the proposed ISAA was useful for evaluating risks within a system containing complicated uncertainties and interactions and providing supports for identifying cost-effective site management strategies.

A factorial-design-based fuzzy-stochastic modeling system (FFSMS) was developed in this study to systematically investigate impacts of uncertainties associated with hydrocarbon contaminant transport in subsurface through integration of a compositional model, factorial design method, fuzzy modeling approach and Monte Carlo simulation technique. The goodness of fit of the numerical model was analyzed by means of a pilot-scale experimental system. Once the model was calibrated, it was used in order to predict the contaminant concentration depending on values of several parameters including intrinsic permeability, porosity, and longitudinal dispersivity. These parameters were imprecisely known, and such an imprecision was handled by means of both fuzzy sets and/or stochastic theory. The individual and joint effects of these uncertain parameters were analyzed by modeling the dependence between the prediction and the imprecise parameters (factors) through factorial design analysis. The study results indicated that the uncertainties associated with input parameters had significant impacts on modeling outputs; the degree of influence of each model input varied significantly with the level of its imprecision. The study results demonstrated that proposed FFSMS can efficiently analyze the impact of different uncertainty sources associated with different hydrogeological parameters on the prediction of the hydrocarbon concentrations in groundwater. Such studies would provide strong basis for performing successful risk assessment and efficient remediation design for the management of contaminated site.

Water quality management is complicated with a variety of uncertainties and nonlinearities. This leads to difficulties in formulating and solving the resulting inexact nonlinear optimization problems. In this study, an inexact chance-constrained quadratic programming (ICCQP) model was developed for stream water quality management. A multi-segment stream water quality (MSWQ) simulation model was provided for establishing the relationship between environmental responses and pollution-control actions. The relationship was described by transformation matrices and vectors that could be used directly in a multi-point-source waste reduction (MWR) optimization model as water-quality constraints. The interval quadratic polynomials were employed to reflect the nonlinearities and uncertainties associated with wastewater treatment costs. Uncertainties associated with the water-quality parameters were projected into the transformation matrices and vectors through Monte Carlo simulation. Uncertainties derived from water quality standards were characterized as random variables with normal probability distributions. The proposed ICCQP model was applied to a water quality management problem in the Changsha section of the Xiangjiang River in China. The results demonstrated that the proposed optimization model could effectively communicate uncertainties into the optimization process, and generate inexact solutions containing a spectrum of wastewater treatment options. Decision alternatives could then be obtained by adjusting different combinations of the decision variables within their solution intervals. Solutions from the ICCQP model could be used to analyze tradeoffs between the wastewater treatment cost and system-failure risk due to inherent uncertainties. The results are valuable for supporting decision makers in seeking cost-effective water management strategies.

An inexact chance-constrained linear programming ICCLP model for optimal water pollution management at the watershed scale was developed. We selected the net expenditures of the alternative strategies, including initial capital investment and operating costs, as the objectives of water pollution management. The total environmental capacity of the water bodies at different probability levels qi was considered a key constraint; other constraints included in the model were government minimum requirements on farmland area, land cover, treatment rate of domestic wastewater and rural wastes, and certain technical constraints. The ICCLP model was applied to Lake Qionghai watershed in China for water quality improvement with the goal of achieving a minimum total cost. Alternative strategies were incorporated following discussions with shareholders and experts. A three-period optimization was conducted based on the alternative strategies; the model parameters were based on field investigations. Five probability levels were considered in the model: qi=0.01, 0.25, 0.50. 0.90, and 0.99. The model results showed that the total optimized costs were between US$55,710.86,80,691.8110 4 and US$72,151.39,101,338.610 4 under different probability levels. The model results indicate that soil erosion treatment, nonpoint source control measures, and rural waste treatment have much higher costs than other strategies, and our findings indicate that the ICCLP model can effectively deal with optimal water pollution management under uncertainty at the watershed scale.

Groundwater contamination by leakage and spill of petroleum hydrocarbons from underground storage tanks has been a major environmental concern. Among various remediation alternatives, the vacuum-enhanced free product recovery (VFPR) is an important technology to extract light nonaqueous-phase liquids (LNAPLs) from subsurface. However, efficient design of a VFPR system was challenging to practitioners, since the process of hydrocarbon removal is costly and time consuming. To address such a problem, an integrated study system for optimizing the VFPR process was developed through coupling a numerical modeling system, a multivariate regression technique and nonlinear optimization model into a general framework. A two-dimensional multiphase flow simulation system was provided for modeling VFPR processes. An iterative stepwise-inference regression (ISIR) method was advanced for establishing a linkage between remediation actions and system responses. A nonlinear optimization model embedded with ISIR was then established for generating desired operating conditions. The results from a case study demonstrated that the established optimization model could effectively analyze tradeoffs between various environmental and economical considerations, and provide effective decision supports for site remediation practices. Compared with the conventional stepwise-cluster analysis method, the proposed ISIR method was more efficient and reliable in approximating relationships between remediation actions and system responses, and could significantly enhance the robustness of optimization solutions.

An integrated fuzzy simulation-assessment method (FSAM) was developed for assessing environmental risks from petroleum hydrocarbon contamination in ground water. In the FSAM, techniques of fuzzy simulation and fuzzy risk assessment were coupled into a general framework to reflect a variety of system uncertainties. A petroleum-contaminated site located in western Canada was selected as a study case for demonstrating applicability of the proposed method. The risk assessment results demonstrated that system uncertainties would significantly impact expressions of risk-level outputs. A relatively deterministic expression of the risks would have clearer representations of the study problem but may miss valuable uncertain information; conversely, an assessment under vaguer system conditions would help reveal potential consequences of adverse effects but would suffer from a higher degree of fuzziness in presenting the modeling outputs. Based on the risk assessment results, a decision analysis procedure was used to calculate a general risk index (GRI) to help identify proper responsive actions. The proposed method was useful for evaluating risks within a system containing multiple factors with complicated uncertainties and interactions and providing support for identifying proper site management strategies.

[1] An integrated simulation-optimization system was developed for supporting decisions of the dual phase vacuum extraction (DPVE) processes. The system coupled a DPVE process simulator, a multivariate regression tool and a nonlinear optimization model into a general framework. A stepwise-cluster-analysis technique was provided for establishing a DPVE process forecasting system for describing the relationships between remediation actions and system responses (i.e., total extracted volume of oil/water, elevation distribution of water table, and specific volume of oil). The forecasting system was then embedded into a multiobjective optimization framework, where the objectives were to minimize the operation cost and maximize the remediation efficiency. The constraints include environmental, hydraulic and technical restrictions to the DPVE processes. A case study was conducted for a petroleum-contaminated site in western Canada. The results from the stepwise cluster analysis indicated that the generated cluster trees could be used for predicting system responses of the DPVE process, given inputs of the operating conditions. The prediction accuracies of the generated cluster trees were verified using randomly generated data sets. The optimum operating conditions could vary significantly under different cost-efficiency targets. When a stricter environmental target (i.e., the amount of pollutants in subsurface) was concerned, a higher system cost had to be paid; when the cost became a critical factor, the performance of contaminant removal would have to be compromised. The developed system could be used to analyze tradeoffs between system cost and process efficiency in the DPVE operations; it could also support the formulation of an on-site process-control system with vacuum levels and extraction rates being the main control variables.