Study On The Law And Prediction Method Of Return Water In Large-scale Yellow River Diversion Irrigation Area

On the basis of consulting relevant domestic and foreign documents and materials,this paper aims at the problems of large return water,many influencing factors and complicated return water laws in China's large-scale Yellow River irrigation areas,and adopts a technical route that combines actual monitoring,theoretical analysis,numerical simulation and empirical research,combined with the actual situation of Gansu Jingdian Irrigation District,mainly carried out the research on the law of return water and the influencing factors,the temporal and spatial distribution characteristics of the return water,the prediction method of the return water volume,the optimal allocation of water resources and the utilization method of the return water in the large-scale Yellow River irrigation area.The main research results are as follows:(1)The regularity of return water in the irrigation area and the dynamic change characteristics of groundwater depth are studied,and the composition of return water,the characteristics of return water,the relationship between return water and influencing factors,and the response of groundwater depth to irrigation are clarified.The return water composition of Jingdian Irrigation District mainly includes deep seepage water generated by irrigation,mountain torrents,underground drainage,surface runoff,and a small part of domestic sewage and industrial waste water.Influenced by irrigation,precipitation,evaporation and the change of groundwater level,the drainage law of irrigation ditches has different characteristics in different spatial scales and different catchment areas.The results show that the annual variation of the first stage irrigation area and the catchment area of Nansha River is small in irrigation month and large in non irrigation month;In hongbiliang catchment area,the annual variation of water recession is larger in summer,but smaller in other months.Among the influencing factors of channel return water in Jingdian irrigation area,groundwater depth and irrigation have a greater impact.Among them,return water and evaporation are negatively correlated,and the changes are consistent,but there is a certain lag with other factors.The variation of groundwater depth in different irrigation areas is regulated by the vadose zone,which has obvious hysteresis relative to irrigation.(2)The hysteresis relationship between the amount of water receding and the influencing factors is analyzed,and a method for predicting the amount of receding water with hysteresis is proposed,which improves the prediction accuracy of the amount of return water.The cross-wavelet analysis method is used to determine the lag time of the return water with respect to the irrigation and precipitation in different spatial scales and different confluence areas.Among them,the lag time of the first stage irrigation area overall return water and Nansha River Xiangshui return water relative to the irrigation amount is longer,but the lag time relative to the precipitation is shorter;Compared with irrigation and precipitation,the lag time of Wufo recession in hongbiliang is short.Based on the hysteresis of the influencing factors of water receding,a neural network model with input elements in staggered periods is proposed to predict the water receding volume.Based on the lag of the regression to the influencing factors,the neural network model with staggered period input factors is proposed to predict the regression.Compared with the same period input,the prediction accuracy of the improved method is significantly improved,and the prediction accuracy of the sub catchment area is higher than that of the first stage irrigation area.At the same time,the way of inputting factors in staggered period is also suitable for other regression prediction models.(3)Aiming at the calculation of the return water volume in areas without return water monitoring data,a method for estimating the return water volume suitable for this condition is proposed,which provides an idea for the calculation of the return water volume.By calculating the water consumption of farmland irrigation,ecology,production and living in the irrigation area,the overall and local water consumption coefficients of the irrigation area are obtained,and then the proportion and the amount of return water between the whole irrigation area and each catchment area are estimated;Among them,the proportion of return water in each subregion of Jingdian Phase ? irrigation area is significantly higher than that of each subregion of Phase ?irrigation area.Through the calculation of water consumption coefficient method,it is revealed that the proportion of water withdrawal is large,the proportion of water consumption in water supply is small,and the utilization efficiency of water resources is low.At the same time,through the calculation of water supply and demand in irrigation area,it shows that the contradiction between water supply and demand in irrigation area is prominent,and the water demand of salt leaching increases the demand of overall water resources in irrigation area.The water consumption coefficient method provides a basis for water resources management in areas without data.(4)The mechanism of return water in the irrigation area was revealed,the distributed coupling model of MIKE SHE and MIKE 11 was constructed,and the amount of return water in the irrigation area was calculated from the perspective of water circulation,and the relationship between the process and transformation of water retreat was clarified.By simulating the hydrological process of each catchment basin in the irrigation area,the characteristics of surface runoff and groundwater level change were revealed,and the main water circulation process in the basin was cleared as water diversion irrigation-deep seepage-underground drainage-surface runoff.Based on the Nash Sutcliffe coefficient and root mean square error evaluation results of the model in the rate period and validation period,the coupling model can better simulate the water cycle process in the irrigation area.Then,the difference between the simulated runoff with and without irrigation was analyzed to get the irrigation return water.From the perspective of water balance analysis,the excess irrigation water is supplied to groundwater through deep infiltration,making the water in the saturated zone of the irrigation area in a surplus state.(5)According to the calculation and simulation of the return water in the irrigation area,the current state of the water resources in the irrigation area is studied,a multi-objective optimization model is constructed,the water and soil resources in the irrigation area are rationally allocated,and the use of the return water is proposed.On the basis of obtaining the current and future water and soil resource parameters,the model aims at maximizing economic benefits,grain output,ecological benefits and water production efficiency.The particle swarm method is used to solve the optimization model.The results showed that the proportion of grain crops decreased and the crop planting structure tended to be more reasonable in the current year and the planning year after optimization;The crop irrigation quota decreased significantly and the comprehensive benefit is the best.By optimizing the irrigation quota,the amount of water returned in the irrigation area is fundamentally reduced.At the same time,according to the characteristics of the amount of water returned,the method of re-irrigation and ecological water use of the returned water is proposed to realize the efficient use of water resources in the irrigation area.