Guanhui Cheng

A Hybrid Dynamic Dual Interval Programming for Irrigation Water Allocation under Uncertainty

Along with the economic development in Canada, the shortage of irrigation water has become a serious concern (Bouwer 1993; Hennessy 1993). In this study, a model of Dynamic Dual Interval Programming (DDIP) is developed and applied to the irrigation water allocation systems with uncertainty. DDIP method improves the existing dynamics interval programming by explicitly addressing the system uncertainties with a dual interval that had higher system reliability. The solution of DDIP is computationally effective, and its decision variables are incorporated into the solutions for final decision. In order to obtain the optimal allocation schemes in a dynamic process, the developed DDIP was applied to an irrigation water system. The results from this case study revealed that optimal solution can be obtained through the DDIP approach from the agriculture water management activities for feasible decisions. These decisions reflect the high uncertainty of the information in the boundaries of dual intervals. The solution presents a maximum benefit under limited yearly uncertain natural resources. Furthermore, the information obtained though this model may help the authority to make optimal decisions and to reduce the risk for uncertain situations.

A stepwise-cluster forecasting approach for monthly streamflows based on climate teleconnections

In this study, a stepwise cluster forecasting (SCF) framework is proposed for monthly streamflow prediction in Xiangxi River, China. The developed SCF method can capture discrete and nonlinear relationships between explanatory and response variables. Cluster trees are generated through the SCF method to reflect complex relationships between independent (i.e. explanatory) and dependent (i.e. response) variables in the hydrologic system without determining specific linear/nonlinear functions. The developed SCF method is applied for monthly streamflow prediction in Xiangxi River based on the local meteorological records as well as some climate index. Comparison among SCF, multiple linear regression, generalized regression neural network, and least square support vector machine methods would be conducted. The results indicate that the SCF method would produce good predictions in both training and testing periods. Besides, the inherent probabilistic characteristics of the SCF predictions are further analyzed. The results obtained by SCF can presented as intervals, formulated by the minimum and maximum predictions as well as the 5 and 95xa0% percentile values of the predictions, which can reflect the variations in streamflow forecasts. Therefore, the developed SCF method can be applied for monthly streamflow prediction in various watersheds with complicated hydrologic processes.

Development of a Stepwise-Clustered Hydrological Inference Model

AbstractFlow prediction is one of the most important issues in modern hydrology. In this study, a statistical tool, stepwise-clustered hydrological inference (SCHI) model, was developed for daily streamflow forecasting. The SCHI model uses cluster trees to represent the nonlinear and complex relationships between streamflow and multiple factors related to climate and watershed conditions. It allows a great deal of flexibility in watershed configuration. The proposed model was applied to the daily streamflow forecasting in the Xiangxi River watershed, China. The correlation coefficient for calibration (1991–1995) was 0.881, and that for validation (1996–1998) was 0.771. Nash–Sutcliffe efficiencies for calibration and validation were 0.768 and 0.577, respectively. The results were compared to those of a conventional process-based model, and it was found that the SCHI model had a superior performance. The results indicate that the proposed model could provide not only reliable and efficient daily flow predic...

Convex contractive interval linear programming for resources and environmental systems management

It is likely that the most reliable estimation of system uncertainty in resources and environmental systems management (RESM) is a value range with an unknown distribution. Stochastic programming would be challenged by distortion of the original uncertain information through fabricating an inexistent probabilistic distribution function. Instead, interval linear programming (ILP), i.e. a synthesis of interval-set coefficients and the conventional linear programming, has been employed to identify the desired schemes for a number of RESM problems under interval uncertainty. However, its effectiveness is disabled by constraint violation which may lead to severe penalties on socio-economic or eco-environmental development. To mitigate such a challenge, a convex contractive interval linear programming (CCILP) approach is proposed in this study. It mainly consists of six modules: parameterizing an RESM problem as an ILP model, initializing a hyperrectangle decision space by two linear programming sub-models, revealing causes of constraint violation given a criterion, inferring feasibilities of potential solutions, finalizing a feasible hyperrectangle decision space by another linear programming sub-model, and supporting RESM of various complexities through alternative variants. A simple ILP model for RESM is introduced to demonstrate the procedures of CCILP and verify its advantages over existing ILP methods. The result indicates that CCILP is capable of robustly incorporating interval uncertainties into the optimization process, avoiding heavy computation burdens on complicated sub-models, eliminating occurrence of constraint violation, enabling provision of a hyperrectangle decision space, adapting to diverse system requirements, and increasing reliability of decision support for interval linear RESM problems.

Synchronic interval Gaussian mixed-integer programming for air quality management.

To reveal the synchronism of interval uncertainties, the tradeoff between system optimality and security, the discreteness of facility-expansion options, the uncertainty of pollutant dispersion processes, and the seasonality of wind features in air quality management (AQM) systems, a synchronic interval Gaussian mixed-integer programming (SIGMIP) approach is proposed in this study. A robust interval Gaussian dispersion model is developed for approaching the pollutant dispersion process under interval uncertainties and seasonal variations. The reflection of synchronic effects of interval uncertainties in the programming objective is enabled through introducing interval functions. The proposition of constraint violation degrees helps quantify the tradeoff between system optimality and constraint violation under interval uncertainties. The overall optimality of system profits of an SIGMIP model is achieved based on the definition of an integrally optimal solution. Integer variables in the SIGMIP model are resolved by the existing cutting-plane method. Combining these efforts leads to an effective algorithm for the SIGMIP model. An application to an AQM problem in a region in Shandong Province, China, reveals that the proposed SIGMIP model can facilitate identifying the desired scheme for AQM. The enhancement of the robustness of optimization exercises may be helpful for increasing the reliability of suggested schemes for AQM under these complexities. The interrelated tradeoffs among control measures, emission sources, flow processes, receptors, influencing factors, and economic and environmental goals are effectively balanced. Interests of many stakeholders are reasonably coordinated. The harmony between economic development and air quality control is enabled. Results also indicate that the constraint violation degree is effective at reflecting the compromise relationship between constraint-violation risks and system optimality under interval uncertainties. This can help decision makers mitigate potential risks, e.g. insufficiency of pollutant treatment capabilities, exceedance of air quality standards, deficiency of pollution control fund, or imbalance of economic or environmental stress, in the process of guiding AQM.

High-resolution projections of mean and extreme precipitations over China through PRECIS under RCPs

The impact of global warming on the characteristics of mean and extreme precipitations over China is investigated by using the Providing REgional Climate Impacts for Studies (PRECIS) model. The PRECIS model was driven by the Hadley Centre Global Environment Model version 2 with Earth System components and coupling (HadGEM2-ES). The results of both models are analyzed in terms of mean precipitation and indices of precipitation extremes (R95p, R99p, SDII, WDF, and CWD) over China at the resolution of 25xa0km under the Representative Concentration Pathways 4.5 and 8.5 (RCP4.5 and RCP8.5) scenarios for the baseline period (1976–2005) and two future periods (2036–2065 and 2070–2099). With improved resolution, the PRECIS model is able to better represent the fine-scale physical process than HadGEM2-ES. It can provide reliable spatial patterns of precipitation and its related extremes with high correlations to observations. Moreover, there is a notable improvement in temporal patterns simulation through the PRECIS model. The PRECIS model better reproduces the regional annual cycle and frequencies of daily precipitation intensity than its driving GCM. Under RCP4.5 and RCP8.5, both the HadGEM2-ES and the precis project increasing annual precipitation over the entire country for two future periods. Precipitation increase in winter is greater than the increase in summer. The results suggest that increased radiative forcing from RCP4.5 to RCP8.5 would further intensify the magnitude of projected precipitation changes by both PRECIS and HadGEM2-ES. For example, some parts of south China with decreased precipitation under RCP4.5 would expect even less precipitation under RCP8.5; regions (northwest, northcentral and northeast China) with increased precipitation under RCP4.5 would expect more precipitation under RCP8.5. Apart from the projected increase in annual total precipitation, the results also suggest that there will be an increase in the days with precipitation higher than 15xa0mm and a decrease in the days with precipitation less than 5xa0mm. Under both RCPs, there would be an increasing trend in the magnitude of changes in precipitation extremes indices (R95p, R99p, and SDII) over China, while an opposite trend is projected for CWD and no apparent trend is projected for WDF from 2036–2065 to 2070–2099. Increased extreme precipitation amounts accompanied with decreased frequencies of extreme precipitation suggest that the future daily extreme precipitation intensity is likely to become large in northeast China and south China.

Dynamically-downscaled temperature and precipitation changes over Saskatchewan using the PRECIS model

In this study, dynamically-downscaled temperature and precipitation changes over Saskatchewan are developed through the Providing Regional Climates for Impacts Studies (PRECIS) model. It can resolve detailed features within GCM grids such as topography, clouds, and land use in Saskatchewan. The PRECIS model is employed to carry out ensemble simulations for projections of temperature and precipitation changes over Saskatchewan. Temperature and precipitation variables at 14 weather stations for the baseline period are first extracted from each model run. Ranges of simulated temperature and precipitation variables are then obtained through combination of maximum and minimum values calculated from the five ensemble runs. The performance of PRECIS ensemble simulations can be evaluated through checking if observations of current temperature at each weather station are within the simulated range. Future climate projections are analyzed over three time slices (i.e., the 2030s, 2050s, and 2080s) to help understand the plausible changes in temperature and precipitation over Saskatchewan in response to global warming. The evaluation results show that the PRECIS ensemble simulations perform very well in terms of capturing the spatial patterns of temperature and precipitation variables. The results of future climate projections over three time slices indicate that there will be an obvious warming trend from the 2030s, to the 2050s, and the 2080s over Saskatchewan. The projected changes of mean temperature over the whole Saskatchewan area is [0, 2]u2009°C in the 2030s at 10th percentile, [2, 5.5]u2009°C in the 2050s at 50th percentile, and [3, 10]u2009°C in the 2090s at 90th percentile. There are no significant changes in the spatial patterns of the projected total precipitation from the 2030s to the end of this century. The minimum change of the projected total precipitation over the whole Province of Saskatchewan is most likely to be −1.3% in thexa02030s, and −0.2% in thexa02050s, while the minimum value would be −2.1% to the end of this century at 50th percentile.

Identification of water quality management policy of watershed system with multiple uncertain interactions using a multi-level-factorial risk-inference-based possibilistic-probabilistic programming approach

In this study, a multi-level-factorial risk-inference-based possibilistic-probabilistic programming (MRPP) method is proposed for supporting water quality management under multiple uncertainties. The MRPP method can handle uncertainties expressed as fuzzy-random-boundary intervals, probability distributions, and interval numbers, and analyze the effects of uncertainties as well as their interactions on modeling outputs. It is applied to plan water quality management in the Xiangxihe watershed. Results reveal that a lower probability of satisfying the objective function (θ) as well as a higher probability of violating environmental constraints (qi) would correspond to a higher system benefit with an increased risk of violating system feasibility. Chemical plants are the major contributors to biological oxygen demand (BOD) and total phosphorus (TP) discharges; total nitrogen (TN) would be mainly discharged by crop farming. It is also discovered that optimistic decision makers should pay more attention to the interactions between chemical plant and water supply, while decision makers who possess a risk-averse attitude would focus on the interactive effect of qi and benefit of water supply. The findings can help enhance the model’s applicability and identify a suitable water quality management policy for environmental sustainability according to the practical situations.

Distributed mixed-integer fuzzy hierarchical programming for municipal solid waste management. Part I: System identification and methodology development

Due to the existence of complexities of heterogeneities, hierarchy, discreteness, and interactions in municipal solid waste management (MSWM) systems such as Beijing, China, a series of socio-economic and eco-environmental problems may emerge or worsen and result in irredeemable damages in the following decades. Meanwhile, existing studies, especially ones focusing on MSWM in Beijing, could hardly reflect these complexities in system simulations and provide reliable decision support for management practices. Thus, a framework of distributed mixed-integer fuzzy hierarchical programming (DMIFHP) is developed in this study for MSWM under these complexities. Beijing is selected as a representative case. The Beijing MSWM system is comprehensively analyzed in many aspects such as socio-economic conditions, natural conditions, spatial heterogeneities, treatment facilities, and system complexities, building a solid foundation for system simulation and optimization. Correspondingly, the MSWM system in Beijing is discretized as 235 grids to reflect spatial heterogeneity. A DMIFHP model which is a nonlinear programming problem is constructed to parameterize the Beijing MSWM system. To enable scientific solving of it, a solution algorithm is proposed based on coupling of fuzzy programming and mixed-integer linear programming. Innovations and advantages of the DMIFHP framework are discussed. The optimal MSWM schemes and mechanism revelations will be discussed in another companion paper due to length limitation.

Distributed mixed-integer fuzzy hierarchical programming for municipal solid waste management. Part II: scheme analysis and mechanism revelation.

As presented in the first companion paper, distributed mixed-integer fuzzy hierarchical programming (DMIFHP) was developed for municipal solid waste management (MSWM) under complexities of heterogeneities, hierarchy, discreteness, and interactions. Beijing was selected as a representative case. This paper focuses on presenting the obtained schemes and the revealed mechanisms of the Beijing MSWM system. The optimal MSWM schemes for Beijing under various solid waste treatment policies and their differences are deliberated. The impacts of facility expansion, hierarchy, and spatial heterogeneities and potential extensions of DMIFHP are also discussed. A few of findings are revealed from the results and a series of comparisons and analyses. For instance, DMIFHP is capable of robustly reflecting these complexities in MSWM systems, especially for Beijing. The optimal MSWM schemes are of fragmented patterns due to the dominant role of the proximity principle in allocating solid waste treatment resources, and they are closely related to regulated ratios of landfilling, incineration, and composting. Communities without significant differences among distances to different types of treatment facilities are more sensitive to these ratios than others. The complexities of hierarchy and heterogeneities pose significant impacts on MSWM practices. Spatial dislocation of MSW generation rates and facility capacities caused by unreasonable planning in the past may result in insufficient utilization of treatment capacities under substantial influences of transportation costs. The problems of unreasonable MSWM planning, e.g., severe imbalance among different technologies and complete vacancy of ten facilities, should be gained deliberation of the public and the municipal or local governments in Beijing. These findings are helpful for gaining insights into MSWM systems under these complexities, mitigating key challenges in the planning of these systems, improving the related management practices, and eliminating potential socio-economic and eco-environmental issues resulting from unreasonable management.