| Affected by climate change,drought in China is frequent and widespread,which seriously affects social and economic stability and development as well as Balance of nature.Utilizing reservoir regulation for drought resistance and disaster reduction is one of the important non engineering measures to cope with drought.Therefore,it is necessary to conduct research on reservoir drought resistance regulation and drought warning water level.This article takes the Jinghui Canal irrigation area as the research area,sets up multiple combined schemes for the destruction of irrigation water supply during crop growth periods,and simulates the disaster situation of crops under different schemes using the AquaCrop-OSPy crop growth model.It studies the agricultural production loss curve due to drought and the utilization efficiency of water resources in agriculture;Taking Dongzhuang Reservoir as the research object,combined with the AquaCrop-OSPy model,a reservoir drought resistance operation model is established with the goal of minimizing agricultural production losses due to drought.The operation status of the reservoir under different operating modes and the water supply situation of the water department are compared and analyzed;By coupling the regional drought Index set and the water resource situation Index set,the drought evaluation Index set is established with the month as the time period.The number of drought evaluation indicator clusters in each month is determined by using the contour coefficient method,and the Index set is clustered by using the K-means clustering algorithm,and the reservoir operation period is staged by combining Fisher optimal segmentation method,so as to determine the hierarchical and staged drought warning level of Dongzhuang Reservoir.The research methods and main results are as follows:(1)Research on the quantification and evaluation of agricultural production losses due to drought.Taking the Jinghui Canal irrigation area as the research area,a combination scheme of multiple crop growth period irrigation water supply disruptions was set up.Based on the calibrated AquaCrop-OSPy model,agricultural production disasters were simulated,and the relationship between crop yield reduction rate and water shortage degree was analyzed.It was determined that the crop loss curve due to drought is a continuous,multi inflection point,threestage change,and monotonically increasing S-shaped curve;Based on variance decomposition to quantify the impact of monthly water scarcity rate on crop yield,the results show that the benefits of agricultural water utilization vary in different months.The water scarcity rate in July has the greatest impact on summer maize,and the water scarcity rate in December has the greatest impact on winter wheat.(2)Research on reservoir drought relief operation.Taking Dongzhuang Reservoir as the research object,three reservoir operation models are established.In model Ⅰ,the reservoir operates according to standardized operation strategy rules,and combined with the AquaCropOSPy model to simulate crop yields;Model Ⅱ and model Ⅲ respectively take the reservoir discharge and reservoir operation line as decision variables,combine with AquaCrop-OSPy model,and take the minimum loss of agricultural output due to drought as the goal,both adopt Particle swarm optimization algorithm to solve,and conduct a long series of simulated dispatching.Comparing and analyzing the operation mode and water supply process of reservoirs in the three models,the results show that Model Ⅱ and Model Ⅲ can effectively utilize the feedback messages from the AquaCrop-OSPy model,which can minimize agricultural losses due to drought and improve water resource utilization while meeting the water needs of daily life,industry,and ecology;In Model Ⅱ,the industrial living guarantee rate is the highest at 96%,followed by Model Ⅲ,and Model Ⅰ has the lowest;The total yield of winter wheat in Model Ⅱand Model Ⅲ increased by 16%and 10%compared to Model Ⅰ,respectively;When the incoming water is relatively abundant,the operation process of the reservoirs in the three models is relatively similar;In normal water years and when the inflow is small,there are differences in the operation process of the reservoir.Among them,Model Ⅱ and Model Ⅲ can improve the reservoir water level and water resource utilization efficiency by restricting water supply,and can effectively carry out drought resistance and disaster reduction.(3)Research on reservoir drought warning and drought warning water level.By summarizing existing research on drought warning water levels,it is concluded that expanding the reservoir drought warning water level refers to the water level warning threshold that,with the formation and development of drought,the balance between water resource supply and demand is about to be broken and the rapid development of drought is triggered.It is the controlled water level at which the basin or regional water resource allocation system enters the stage of restricted water supply;At the same time,the drought warning water level has grading attributes,staging attributes,hedging attributes,critical attributes,and planning attributes;The Index set for regional drought assessment is established by using the water storage at the beginning of the month and soil moisture content in Dongzhuang Reservoir in Model Ⅲ.The Index set for water resource situation assessment is taken as the inflow runoff and regional rainfall of the reservoir,which is coupled to the Index set for reservoir drought assessment.After standardizing the Index set,the number of clusters classified in each month is determined by using the contour coefficient method,and then the Index set is cluster analyzed by using the Kmeans algorithm,Using the lowest water level within various clusters as the graded drought warning water level for each month;Using Fisher’s optimal segmentation method,the reservoir operation period is divided into three periods:July to October,November to December,and January to June;Finally,the graded and staged drought alarm water levels of Dongzhuang Reservoir will be determined by merging and scheduling the monthly graded drought alarm water levels within each stage. |