Expert knowledge based modeling for integrated water resources planning and management in the Zayandehrud River Basin
Expert knowledge based modeling for integrated water resources planning and management in the Zayandehrud River Basin
复制标题
DOI：
10.1016/j.jhydrol.2015.07.014
复制
发表时间：
2015-09
影响因子：
6.4
通讯作者：
H. Safavi;M. Golmohammadi;S. Sandoval‐Solis
中科院分区：
文献类型：
--
作者：
H. Safavi;M. Golmohammadi;S. Sandoval‐Solis
这项研究强调了水资源规划和管理的需要，利用专家知识来模拟复杂的水文系统中已知的极端水文变异性，缺乏数据。伊朗的Zayandehrud河流域被用作复杂水系统的一个例子;这项研究提供了该流域的全面描述，包括其水需求（市政，农业，工业和环境）和供水资源（河流，跨流域调水和含水层）。本研究的目的是评估该盆地近期的条件（从10月/ 2015年9月/ 2019年），考虑到当前的水资源管理政策和气候变化条件，称为Baselinescenario。建立了Zayandehrud流域的规划模型，以评估基线情景，水文分析期为21年（从10月/ 1991年至9月2011年）;它被校准了17年，并验证了4年使用aHistoricscenario，考虑到历史供水，基础设施和水文条件。由于TheZayandehrud模型是一个规划模型，而不是一个水文模型（径流模型），自适应网络模糊推理系统（ANFIS）是用来生成合成的自然流量考虑温度和降水作为输入。该模型是一个基于专家知识和数据的模型，具有人工神经网络（ANN）和模糊推理系统（FIS）的优点。ANFIS模型的输出与历史情景结果进行了比较，并用于基线情景。使用三个度量来评估ANFIS模型的拟合优度。使用五个性能标准：基于时间和体积的可靠性，弹性，脆弱性和最大赤字的基线方案的供水结果进行了分析。其中一个指数，即水资源可持续性指数，被用来总结业绩标准的结果，并便于比较各种利弊。基线情景的结果表明，水需求的供应将以消耗地表水和地下水资源为代价，使这种情景不可取，不可持续，并有可能对水源造成不可逆转的负面影响。因此，目前的水资源管理政策是不可行的;需要制定更多的水资源管理政策，通过提高灌溉效率和减少地下水开采，减少对水的需求，以实现Zayandehrud盆地的可持续水资源管理。
This study highlights the need for water resource planning and management using expert knowledge to model known extreme hydrologic variability in complex hydrologic systems with lack of data. The Zayandehrud River Basin in Iran is used as an example of complex water system; this study provides a comprehensive description of the basin, including its water demands (municipal, agricultural, industrial and environmental) and water supply resources (rivers, inter-basin water transfer and aquifers). The objective of this study is to evaluate near future conditions of the basin (from Oct./2015 to Sep./2019) considering the current water management policies and climate change conditions, referred asBaselinescenario. A planning model for the Zayandehrud basin was built to evaluate the Baseline scenario, the period of hydrologic analysis is 21 years, (from Oct./1991 to Sep./2011); it was calibrated for 17 years and validated for 4 years using aHistoricscenario that considered historic water supply, infrastructure and hydrologic conditions. Because theZayandehrud modelis a planning model and not a hydrologic model (rainfall–runoff model), an Adaptive Network-based Fuzzy Inference System (ANFIS) is used to generate synthetic natural flows considering temperature and precipitation as inputs. This model is an expert knowledge and data based model which has the benefits of Artificial Neural Networks (ANN) and Fuzzy Inference Systems (FIS). Outputs of the ANFIS model were compared to the Historic scenario results and are used in the Baseline scenario. Three metrics are used to evaluate the goodness of fit of the ANFIS model. Water supply results of the Baseline scenario are analyzed using five performance criteria: time-based and volumetric reliability, resilience, vulnerability and maximum deficit. One index, the Water Resources Sustainability Index is used to summarize the performance criteria results and to facilitate comparison among trade-offs. Results for the Baseline scenario show that water demands will be supplied at the cost of depletion of surface and groundwater resource, making this scenario undesirable, unsustainable and with the potential of irreversible negative impact in water sources. Hence, the current water management policy is not viable; there is a need for additional water management policies that reduce water demand through improving irrigation efficiency and reduction of groundwater extraction for sustainable water resources management in the Zayandehrud basin.