Development Of Decision Support System For Irrigation And Fertilization In Vineyard

Grapes are a characteristic economic fruit that has been vigorously developed in the Guanzhong region of Shaanxi in recent years.Since the growth and development period of grapes is in the period of high frequency of drought occurrence,drought has become a major limiting factor to improve the productivity of grapes.In the context of climate change,the adoption of only traditional static water allocation and irrigation system schemes may result in crop yield reduction and water wastage.Therefore,on the basis of analyzing the changes of water demand and water surplus and deficit of grapes under climate change conditions,we developed a vineyard irrigation and fertilization decision support system using Python as the programming language.This study provides theoretical basis and technical support for scientific formulation of irrigation strategies and improvement of water resource utilization efficiency in the Guanzhong Plain,and is important for changing the traditional irrigation methods in China,realizing agricultural information and intelligent management,and serving rural revitalization.(1)The multi-year average relative humidity and sunshine hours showed a decreasing trend with interannual variation,while the average temperature and wind speed showed an increasing trend with interannual variation during the whole grapevine reproductive period.The water demand of grapes showed a significant increase after the abrupt change in 2015,with 15 a being the first main cycle of water demand variation.The crop water demand tended to increase and then decrease during the grapevine reproductive period,and the difference in water demand between hydrological years reached its maximum at the fruit expansion stage.The water crop water surplus deficit index of grapes during the period 1990-2019 was relatively flat,with a mean value of-26.22%.The risk of drought was different in all hydrological years,and special attention should be paid to the water deficit of grapes in June in wet and medium years to ensure adequate water supply during the critical period of grape water demand.(2)The vineyard irrigation and fertilization decision support system is developed based on Windows 10 operating system,Python programming language,and Py Charm integrated development environment.The software uses the Django framework and the architecture uses the MVC(Model View Controller)mode.Based on the overall functional requirements,the system was developed in three modules:the parameter acquisition and data processing module,the grape irrigation and fertilization decision support module,and the irrigation auxiliary decision module.(3)The parameter acquisition and data processing module includes the user input parameters,the acquisition and screening of rainfall and Bowen-ratio observation system data.After obtaining the parameters,the ORM framework is used to read and write the target database through the model.The water requirement for grape irrigation is calculated using the Bowen-ratio energy balance method,water balance method,and the system provides fertilizer application recommendations for different fertility periods.The irrigation assisted decision module includes the irrigation assisted decision based on the soil moisture model of maximum daily shrinkage of stalk and the irrigation assisted decision based on the weather forecast,in which the irrigation assisted decision based on the soil moisture model of maximum shrinkage of stalk can improve the decision accuracy of the system,and the irrigation assisted decision based on the weather forecast can save irrigation water while ensuring the water requirement of grapes.(4)A simple,beautiful and interactive system interface was developed.The overall design is user-driven and easy for users to understand and operate.In terms of system testing,the system functions have been basically implemented and initially passed the functional test,ease of use test,and performance test.The coefficient of determination(R~2)between the simulated soil moisture content of the system and the actual soil moisture content without correction is 0.869,the root mean square error(RMSE)is 0.007,the consistency coefficient(d)is 0.954,and the mean absolute error(MAE)is 0.006,the accuracy of the system is good.