Case study on an integrated operation planning of multiple reservoirs for irrigation in Japan

In spite of several attempts at integrated operation planning, multiple reservoirs in Japan have been operated by trial and error without any formal rules. Subjects of integrated operation are not only showing optimal usage of daily storage levels as an operational policy but also providing a countermeasure for droughts.Objectives of this study are to make a formal operation rule of multiple reservoirs for irrigation using the theory of Required Storage for Drought Curve (RSDC) Method and to propose operational policy for multiple reservoirs as large water supply systems. The Iwaki river basin, on which there are four reservoirs for irrigation parallel with each other, is considered to be a model river basin with a large water system for this case study.From results of simulations using historical data, comparing an individual operation rule with an integrated operation rule on several indices, effectiveness of the latter rule is recognized. Under integrated operation, water losses are minimized and excessive water conservation can be avoided over the whole area to benefit while target river discharge is maintained at key control points because water usage from all reservoirs is well balanced in relation to water availability.     

Modelling scenarios to identify a combined sediment-water management strategy for the large reservoirs of the Tuyamuyun hydro-complex

This paper introduces a combined modelling approach using a simple water budget model (THC-model) and a 3D reservoir sedimentation model (MOHID Water) to adapt reservoir operation and visualise their effects on the sediment deposition. By this, an effective combined sediment-water management can be identified under semi-arid conditions for dry, median and wet years. Results are presented for the reservoirs of the Tuyamuyun Hydro-Complex (THC), which is located in the lower Amu Darya River. The determination of the actual and usable reservoir storage volume shows that siltation will significantly adversely affect the ability of the in-stream Channel Reservoir to regulate seasonal demand for both irrigation and municipal water supply. However, modelling scenarios results confirm the effectiveness of adapted operation rules for the THC reservoirs and show that the operation of large dams could be modified according to a combined sediment-water management. The experience gained during this study emphasizes the fact that the concept of a combined reservoir management of sediments as well as water can be an efficient measure to improve the sustainable long-term use of reservoirs and to contribute towards a safe water supply in water crisis regions.     

Planning and management of a complex water resource system: case of Samanalawewa and Udawalawe reservoirs in the Walawe river,Sri Lanka

The Samanalawewa and Udawalawe reservoirs were built to harness the hydro-energy and irrigation potentials of the Walawe river in Sri Lanka. The recently completed Samanalawewa reservoir primarily generates hydropower while the Udawalawe reservoir, which was built in the 1960s, supplies water mainly for irrigation. With the addition of the Samanalawewa reservoir, located upstream of the Udawalawe reservoir, the Government of Sri Lanka is planning to increase the irrigated area of the Udawalawe reservoir. The Samanalawewa reservoir is expected to act as an additional storage for irrigation water supply. A study was carried out to investigate the operational behavior of these two reservoirs. The model used in the study is based on stochastic dynamic programming (SDP) and simulation techniques. Since, the direct application of SDP for two reservoirs is limited by the dimensionality of the problem, a sequential decomposition method is employed in the model. The algorithm employed breaks down the system into single-reservoir subsystems and subsequently, each subsystem’s operation is individually optimized using a SDP based optimization model and then simulated using a reservoir operation simulation model. The results indicate the usefulness of optimization techniques in planning reservoirs and deriving operational policies for them. The inclusion of the Samanalawewa reservoir reduces the irrigation water supply deficits at the Udawalawe reservoir.     

Optimal irrigation planning model for an existing storage based irrigation system in India

A weekly irrigation planning LP model is formulated for determining the optimal cropping pattern and reservoir water allocation for an existing storage based irrigation system in India. Objective of the model is maximization of net annual benefit from the project. In an irrigation planning of a storage based irrigation system, initial storage of the reservoir at the beginning of the reservoir operation, expected inflows into the reservoir during each intraseasonal period, capacity of channels, crop calendar and yield response to water deficit in each growth stage of crop play a vital role in deciding acreage and water allocation to each crop. The planning model takes into account yield response to water deficit in each intraseasonal period of the crop, expected weekly inflows entering into the reservoir, storage continuity of reservoir, land and water availability, equity of water allocation among sub areas and proportionate downstream river release. One year comprising of 52 weeks is considered as planning horizon. To account for uncertainty in water resources availability, the model is solved for four levels of reliability of weekly inflows entering into the reservoir (90%, 85%, 80% and 75%). Alternative optimal cropping patterns and weekly releases to crops grown in each sub area under each main canal are obtained for various states of initial storage at the beginning of reservoir operation and for various levels of weekly inflows into the reservoir. Results reveal the importance of initial state of reservoir storage for feasible solution and shows the impact on cropping pattern with the change in initial storage of reservoir for different levels of reliability of weekly inflows.     

多年调节水库年末消落水位的确定

多年调节水库年末消落水位的合理控制是保证水库(群)长期正常运用的关键。在分析了多年调节水库发电量与水库蓄能关系的基础上建立了确定其年末消落水位的多目标决策模型，并使用模糊多目标动态规划的方法进行求解，在给定权重下可一次性得到多年调节水库的最优年末水位及相应的最优发电策略，为最终决策提供依据。     

平原(灌区)地下水库的数值模拟

在北方平原地区修建浅层地下水库 ,利用浅层含水层来调蓄雨水和跨流域调水 ,有重要的现实意义。分析了灌区用水和地下水库运用的要求 ,提出了浅层地下水库设计库容的确定原则 ,介绍了地下水库数值模拟方法。实例分析表明 :数值模拟可为地下水库提供设计参数及运行方案。     

疏勒河中游梯级水库库容和库水位关系模型优化

为了研究梯级水库库容与库水位之间的关系，解决水库由原来承担单一灌溉或发电任务，到现在变为一库多用的多目标决策问题。以疏勒河流域中游地区的3座水库（昌马水库、双塔水库和赤金峡水库）为研究对象，利用高斯函数和两种神经网络建立3大水库的库水位和库容之间的关系模型。通过对比不同模型预测的均方根误差（RMSE），相对均方根误差（RRMSE）和平均绝对百分比误差（MAPE）来判断模型的表现。结果表明:3种模型均可较好地拟合3大水库库水位和库容之间的关系，对于昌马水库和双塔水库来说，径向基函数神经网络（RBF）的表现优于非线性回归模型和前馈反向传播神经网络（FBNN）模型；对于赤金峡水库而言，FBNN神经网络的表现优于非线性回归和RBF神经网络模型。与经典回归模型相比，人工神经网络更适合于水库库容的确定。      

石津灌区储水灌溉条件下土壤水分动态及灌水适宜阈值研究

从土壤水分能态角度,研究储水灌溉条件下土壤水分的动态变化及空间分布,探求适宜储水灌溉定额阈值范围。研究结果表明,灌水定额大于200 mm时,2 m以下土层出现水分深层渗漏,灌水定额300、2502、00 mm时,深层渗漏量分别达到587.63、236.32、152.05 m3/hm2;灌水定额75～150 mm,2 m以下土层无水分渗漏。因此,储水灌溉灌水定额阈值范围控制在750～1 500 m3/hm2,可以把灌溉水储存于深层土体内,以供作物生长期使用;储水灌溉模式在石津灌区可有效解决灌区来水与灌溉用水的错位矛盾,满足作物正常生长对水分的需求。     

考虑资料缺失的小型水库影响的洪水预报方法研究

流域中修建的大量小型水库会改变洪水的自然特性,而小型水库资料往往缺失或信息不全,使得在洪水预报中难以考虑其调蓄作用,给洪水预报带来困难。利用遥感RS、地理信息系统GIS和支持向量回归机SVR算法,构建地形地貌参数-水库库容关系模型,推算水库资料缺失地区的小型水库库容信息;在定量概化小型水库对洪水拦蓄影响的基础上,建立了新安江-水库模型。以淮河上游北庙集流域为对象进行应用研究,结果表明:(1)小型水库总库容与地形地貌特征有较强的相关关系,建立的地形地貌参数-水库库容模型具有较高精度,可用于资料缺失的小型水库的库容推算;(2)加入水库拦蓄模块后,新安江模型的洪水预报精度得到一定提高。研究成果可供受小型水利工程影响地区的实时洪水预报参考。     

灌溉水库兴利调度初探

该文分析了灌溉水库兴利调度的特点,提出了灌溉水库分干旱类型制定调度线和破坏年调度问题,分析了灌溉水库担任综合供水任务后的调度处理,提出了灌溉水库调度的检验原则。     

龙羊峡节水增发电考核系统研究与

黄河上游梯级调度的原则是龙羊峡和刘家峡两库联合调度，龙羊峡是多年调节水库，针对这一特点，结合龙羊峡和刘家峡两库联调原则，提出多年调节水电站节水增发电考核计算方法，以提高电厂的水能利用率，并开发了相应的可视化计算软件，便于电网调度。     

Integrated decision making for reservoir,irrigation, and crop management

An integrated approach to reservoir, irrigation, and cropping management which links four different models—a hydrologic model (PRMS), a crop growth simulation model (EPIC), an economic model based on linear programming, and a dynamic programming model—is developed and demonstrated. The demonstration is based on an irrigation district located in a subhumid climate with an irrigation reservoir large enough for over-year storage. The model is used to make repeated simulations for various planning horizons. Two different types of results are presented. The first provides the probability that each of the various farm plans (land/crop/water allocation) will be chosen as the optimum in the first year of the planning horizon. The second approach provides probability distributions of accumulated revenues over a chosen length of planning horizon. Each distribution is associated with an initial reservoir level and a particular farm plan in the first year of the planning horizon. The consequence of selecting certain farm plans at the beginning of a specified planning horizon is therefore quantified in a probabilistic way. Based on families of probability–revenue curves, an irrigation manager can simultaneously evaluate crop, irrigation, and reservoir management options.     

Mapping agricultural responses to water supply shocks in large irrigation systems, southern India

Irrigated agriculture experienced a water supply shock during a drought in southern India in 2002-2003. In this paper, hotspots of agricultural change were mapped and put in the context of hydrology and water management. Time series of MODIS imagery taken every eight days before (2001-2002) and during (2002-2003) the supply shock were combined with agricultural census data to document changes in cropping patterns in four large irrigation projects in the downstream sections of the Krishna and Godavari River basins (total command area 18,287 km²). The area cropped in rice in the four irrigated command areas decreased by 32% during the drought year, and rice production in the two districts that experienced the largest flow reductions fell below production levels of 1980. The irrigation project that showed the largest change in double cropped area (−90%) was upstream of the Krishna Delta. In the Krishna Delta, large areas changed from rice-rice to rice-gram double cropping. Historical water management contributed to the vulnerability of rice production to drought: the main reservoir in the system was drained to dead storage levels by the end of each growing season over 1968-2000, with little carryover storage. The land cover change maps suggested that the lower Krishna Basin has experienced a “hard landing” during basin closure, and revised management strategies that account for the new flow regime will be required to maintain agricultural production during droughts.     

Conjunctive use modeling for multicrop irrigation

A mathematical model is developed to arrive at an optimal conjunctive use policy for irrigation of multiple crops in a reservoir-canal–aquifer system. The integration of the reservoir operation for canal release, ground water pumping and crop water allocations during different periods of crop season (intraseasonal periods) is achieved through the objective of maximizing the sum of relative yields of crops over a year considering three sets of constraints: mass balance at the reservoir, soil moisture balance for individual crops, and governing equations for ground water flow. The conjunctive use model is formulated with these constraints linked together by appropriate additional constraints as a deterministic linear programming model. A two-dimensional isotropic, homogeneous unconfined aquifer is considered for modeling. The aquifer response is modeled through the use of a finite element ground water model. A conjunctive use policy is defined by specifying the ratio of the annual allocation of surface water to that of ground water pumping at the crop level for the entire irrigated area. A conjunctive use policy is termed stable when the policy results in a negligible change in the ground water storage over a normal year. The applicability of the model is demonstrated through a case study of an existing reservoir command area in Chitradurga district, Karnataka State, India.     

灌溉渠系运行计算机模拟技术的开发

借鉴日本以及其他国家在灌溉渠系水管理方面的成果和经验，结合我国灌区的实际情况，建立了较完整的渠系运行模型，编制了具有一定通用性和可扩充性的计算机模拟软件。实践表明，该软件对于测试和评价渠系的力学特性、工程控制特性和管理调度特性是有效的，它为改进灌区水管理提供了一个科学、简便、可行的技术手段。     

Adapting to intersectoral transfers in the Zhanghe Irrigation System, China: Part II: Impacts of in-system storage on water balance and productivity

This paper investigates the impacts of farm ponds in a context of declining supplies in a major canal command within the Zhanghe Irrigation System (ZIS), in Central China. As dam supplies have been diverted to higher-valued uses (hydropower, cities and industry), farmers have responded by constructing small storages within their fields. These farm ponds have given them sufficient flexibility in water supply to practice varying forms of alternate wetting and drying irrigation for rice without compromising yields and incomes. Ponds are recharged by a combination of return flows from irrigation and runoff from catchment areas within the irrigated perimeter. Various scenarios of water supply incorporating the main reservoir, in-system reservoirs, farm ponds and irrigation practices were simulated using the OASIS model. OASIS integrates surface and groundwater flows, and contains a crop growth module to aggregate the impacts of different water management regimes. The modelling and sensitivity analysis show that further reductions in main reservoir supplies will have a negative effect on rice production in dry and average years, and that ponds have played a crucial role in adapting agriculture to reduced canal supplies. The flexibility allowed by the ponds has resulted in increased water productivity, except in high rainfall years, but net depletion has not decreased, as local supplies have substituted for water from the main reservoir. The study demonstrates the importance of properly accounting for return flows and the necessity to understand crop production in relation to the actual depletion of water (as evapotranspiration) within an irrigation system.     

Strategies to improve manual operation of irrigation systems in Sri Lanka

Improvements to normal operational procedures are investigated for the main types of irrigation system identified in Sri Lanka. Scheduled and unscheduled changes of flow regime the systems are likely to experience and the opportunities for improved operations related to the physical characteristics of each type are analyzed.Performance of several alternative manual operational procedures are evaluated and compared using a hydraulic model (Simulation of Irrigation Canal, SIC) for three different irrigation subsystem types (Single Bank canal; Double Bank canal; Canals with intermediate Storage). Results of these simulations show that current practice, based on a fixed frequency of operation of 12 hrs and a target water depth set to the spill level of the cross-regulator, is not far from the optimal for systems without on-line storage. A slight improvement can be expected if the frequency of operation is increased to 6 hrs. For systems with on-line storage it is recommended to split operational procedures into two parts, one related to the canal and one to the storage. Among different options investigated, discharge control at the downstream regulator of the storage appears to be the most reliable as it performs well whatever operation procedures are selected for the canal.Irrigation systems in Sri Lanka are subject to frequent positive flow changes (Supply fluctuations, Runoff, Return-flow changes). As far as water management is concerned, improvements to be expected from normal operation procedures appear to be somewhat limited because management of flow changes is not an objective. For instance little effort is made to harvest runoff under the current practices. One conclusion of the study is that new strategies for the harvest, storage and release of positive flow changes should be tested and, where appropriate, implemented to improve the overall efficiency of operation in irrigation systems in Sri Lanka.     

Water ‘banking’ in Fergana valley aquifers—A solution to water allocation in the Syrdarya river basin?

The Syrdarya river is an example of a transboundary basin with contradictory water use requirements between its upstream and downstream parts. Since the winter of 1992-93, the operational regime of the upstream Toktogul reservoir on the Naryn river - the main tributary of the Syrdarya - has shifted from irrigation to hydropower generation mode. This significantly increased winter flow and reduced summer flow downstream of the reservoir. Consequently, excessive winter flow is diverted to the saline depression called Arnasai, while water for summer irrigation is lacking. This study suggests to store the excessive winter flows temporarily in the upstream aquifers of the Fergana valley and to use it subsequently for irrigation in summer. It is estimated that groundwater development for irrigation could be practiced on one-third of the irrigated land of the valley, and conjunctive use of groundwater and canal water on another third; the rest will remain under canal irrigation. This strategy will lower the groundwater table and create aquifer capacity for temporal storage of excessive water—“water banking”. This use of the term is only one of many concepts to which “water banking” or “groundwater banking” is applied. In this paper, the term is applied for temporary storing of river flow in subsurface aquifers. Pilot modeling studies for the Sokh aquifer - one of the 18 aquifers of the Fergana valley - supported that this strategy is a feasible solution for the upstream-downstream issues in the Syrdarya river basin. Field studies of water banking are required to determine the scale of adoption of the proposed strategy for each aquifer of the Fergana valley.     

利用动态规划法确定水库最优灌溉面积

利用动态规划法对供水水库进行长系列优化调度，建立了一维或多维并联供水水库优化调度数学模型，编制了通用计算机程序，通过点绘水库灌溉面积与供水保证率、灌溉面积与缺水率曲线，在给定灌溉保证率下通过查曲线即可确定水训最优灌溉面积。     

A fuzzy risk approach for performance evaluation of an irrigation reservoir system

In this paper, a model for fuzzy risk of low yield of a crop is developed to study the implications of a reservoir operating policy model. When an optimal operating policy is derived based on a known objective, the policy itself does not, in general, indicate a measure of the system performance unless a criterion to this effect is embedded in the objective function. While a systems analyst is interested in the nature of the objective function used in arriving at a policy, the irrigation decision maker would look for the implications of using the policy through answers to the questions such as, how often the system will fail and how quickly it will recover from a failure. It is, therefore, important that the implications of reservoir operation with a given policy be studied keeping in view the interests of the decision makers. Some earlier studies on reservoir operation models for irrigation have considered reliability, resiliency and productivity index, as the performance indicators of the operating policy. In this paper, fuzzy risk of low yield of a crop is considered as another performance indicator to address uncertainties due to both randomness and fuzziness. Uncertainty due to randomness arises primarily because of the random variations of hydrologic variables such as reservoir inflows and rainfall in the command area. Uncertainty due to imprecision or fuzziness arises because of uncertain crop yield response to various factors (such as farm practices and climatic variables) other than to the applied water. Two important concepts are introduced in this paper with respect to irrigation reservoir system. The first one is related to viewing the low yield of a crop, as a fuzzy event. The second concept is related to the definition of fuzzy risk of low yield of a crop. The fuzzy risk of low yield is derived using the concept of probability of a fuzzy event. Application of the methodology is demonstrated with a case study in India.