Font Size: a A A

Methods And Models For Agricultural Water-land Resources Multi-scale Optimal Allocation Based On Uncertainty Analysis

Posted on:2018-02-22Degree:DoctorType:Dissertation
Country:ChinaCandidate:M LiFull Text:PDF
GTID:1319330515482240Subject:Hydraulic engineering
Abstract/Summary:PDF Full Text Request
Developing the water-saving high-efficiency agriculture is the strategy of agricultural sustainable development,and rational allocation of agricultural water-land resources is an important way in improving the efficiency of agricultural resources utilization.Agricultural water-land resources optimal allocation in different special scales have different characteristics.Meanwhile,climate change and human activities result in the uncertainties of agricultural water-land resources optimal allocation systems,leading to the significance to study the agricultural water-land resources optimal allocation methods and models under uncertainty,which will improve agricultural sustainable development under changing environments.Taking the oasis in the middle reaches of Heihe River basin as the study area and on the basis of the clarity of the interactions among different scales of agricultural water-land resources,this study established agricultural water-land resources optimal allocation models under uncertainty and explored the corresponding solutions considering the multiple uncertainties in agricultural water-land resources optimal allocation systems.The contents and conclusions include:(1)Water-land resources allocation in regional scale.Considering the randomness of water supply,a two-level linear fractional programming model was developed and interactive fuzzy algorithm was adopted to solve the model.This model is capable of balancing the benefit between decision makers from different levels,quantitatively solving the ratio multi-objective problems,e.g.,maximum benefit and minimum water use in order to increase system efficiency.Results show that the optimal total water allocation amount of agricultural,industrial,domestic and ecological sectors is 0.24×108m3 less than status quo,and agricultural irrigation proportion of high,middle and low flow levels are 87.6%,84.2%and 82.4%,respectively.The optimal land use for different crops has been adjusted compared to actual conditions under the same total land use,making the total benefit increase by 3.2×108 RMB.(2)Water-land resources allocation in irrigation area scale and risk analysis.A water optimal allocation model based on stochastic simulation of water supply-demand and a planting configuration optimization model based on random fuzzy variable were established.The framework of both the two models is inexact two-stage stochastic programming model,with the characteristic of obtaining the maximum benefit with the minimum water shortage when random events happen,balancing contradictions between gains and losses from resources distribution.In addition,the water allocation model can fully reflect the stochastic characteristic of input parameters,acquiring the optimal water allocation amount with probability distributions.The land allocation model can quantificatively handle dual uncertainties of randomness and fuzziness of input parameters.Result show that the optimal total water allocation amount of the irrigation district alongside the mainstream of Heihe River is 0.54×108m3 less than actual conditions with water allocation efficiency increases,and the optimal total land decreases by 3.17×108ha compared with actual conditions while the per unit benefit increases by 1.28 RMB/ha.Results of evaluation of agricultural irrigation water shortage risks indicate that the water shortage risks in the middle reaches of Heihe River basin are in the category of acceptable risk level or brink risk level.(3)Water-land resources allocation in canal system scale.Taking Yinke irrigation district as a typical example,a flow quantity and time allocation model for main and branch canals,a rotation irrigation group division model for lateral canals,and a crop planting configuration model were established.These models focus on decreasing the leakage loss and increasing allocation efficiency of water-land resources.Results show that the flow quantity of the first round in summer,the second and three rounds in summer,and the first and second rounds in autumn of the second lateral canal of Ymgke irrigation district are 3.32 m3/s,3.42 m3/s and 3.38 m3/s,respectively.The optimal flow quantities in the whole rotation period present uniform,which is beneficial to project implementation in an irrigation district.(4)Water resources allocation in filed scale.An inexact linear fractional programming based on interval crop water production function was established.This model is capable of balancing the contradiction between benefit and water-saving considering the interval uncertainties,with the maximum water productivity obtained.On this basis,in order to coordinate the conflict of interest between decision makers and farmers,a linear fractional-quadratic bi-level programming model for water allocation was established by considering the levels of irrigation districts and crops as a whole.The solving method combining the Lagrange function structure and Kuhn tucker conditions was adopted.Results show that the water productivity of the inexact linear fractional programming model is 1.25 kg/m3 higher than the corresponding values of ordinary linear programing,and the water productivity of linear fractional-quadratic bi-level programming model increases by 4%compared to the water allocation model that is based on quadratic water production function.(5)Agricultural water-land resources planning.Statistical approach combined with intelligence algorithm were used to simulate and forecast the hydrological and social-economic elements.The integration method of fuzzy distinguishing model and cusp catastrophic model was developed to calculate the security thresholds of agricultural water use for different administrative regions,and the combined method of marginal benefit theory and water-saving potential theory was developed to estimate the water use thresholds intervals for the 17 irrigation districts in the middle reaches of Heihe River basin.On these basis,through synthesizing the modelling thoughts of different scales,water-land resources of the whole middle reaches of Heihe River basin were planned and agricultural water use security was evaluated,then a decision support system for the water-land resources allocation was constructed,which will provide reference for the systematic and automated management of the agricultural water-land resources in the middle reaches of Heihe River basin.
Keywords/Search Tags:agricultural water-land resources, optimal allocation, method and model, uncertainty, multi-scale
PDF Full Text Request
Related items