| The Jiansanjiang Administration of Heilongjiang Province is located in the hinterland of the Sanjiang Plain in the north of China.The land within the region is concentrated,and the water and soil resources are abundant.It is suitable for the development of agricultural production.It is an important commodity grain production and reserve base in China.With the continuous development of the economy and population,the demand for food is increasing,and farmers over-exploit land and apply a large amount of chemical fertilizers and pesticides to pursue high yields,so that the health and stability of their agricultural water resources systems are increasingly disturbed,thereby causing groundwater.A series of ecological and environmental problems such as sudden drop in water level,serious water pollution,a sharp drop in water resources,and an increase in soil erosion area has caused the agricultural water resources system to be overloaded.In view of the problems existing in the water resources system of the Sanjiang Administration Bureau mentioned above,this study takes 15 farms under the jurisdiction of the Jiansanjiang Administration of Heilongjiang Province as an example to establish a reasonable regional agricultural water resources system evaluation index system by establishing an index optimization model.The evaluation model calculates the resilience index of the agricultural water resources system,selects the optimal measure method of the resilience of the water resources system by using specific methods,and analyzes the temporal and spatial evolution pattern of the resilience of the agricultural water resources system.Based on the above analysis results,different resilience background constraints are established.The agricultural planting str ucture optimization model provides scientific and technical theoretical guidance for ensuring agricultural water security,food production safety and ecological security.The research results are as follows:(1)This paper analyzes the connotation of the resilience of agricultural water resources system,and selects the preliminary selection of 50 evaluation indicators covering water resources system,agricultural system,socio-economic system and ecological environment system.The basic principle of cluster analysis and maximum information content is the preferred index model.A total of 17 indicators are selected,and finally the regional agricultural water resources system evaluation index system is constructed.The index system established by 34% of the evaluation indicators can reflect the original information of the indicators very reasonably,and referring to the previous research results,from the construction of the index system,the selection of initial indicators and the difference of screening indicators,we can conclude that the index system constructed in this paper is more reasonable and reliable.Based on the Lagrange extreme value combination weighting model,the weights obtained by subjective(Analytic Hierarchy Process)and objective(CRITIC)are synthesized reasonably,and the weighting coefficients of each optimal index are obtained.It can determine the quantified value of each index in the overall proportion very appropriately.(2)According to the established agricultural water resources system resilience index evaluation system,taking the agricultural water resources system of 15 farms in Heilongjiang Province Jiansanjiang Administration in 2015 as an example,using four methods of variable fuzzy set theory,technique for order preference by similarity to an ideal solution(TOPSIS),support vector machine(SVM)and support vector machine based on particle swarm optimization(PSO-SVM)are used to measure the resilience of water resources.According to the serial number sum theory and the Spearman correlation coefficient,the order of each method is calculated,and the stability of the measurement model is obtained.The regenerative level of 15 farms is determined by the Arc GIS natural breakpoint method based on the principle of maximum rank difference.The resilience level reliability of the four evaluation methods is analyzed by using the degree of discrimination theory.Using Arc GIS spatial data analysis technology,spatial distribution map of agricultural water resources system resilience of Jiangsanjiang Administration in 2015 Based on optimal measure model PSO-SVM model is drawn.The results show that the model stability from good to bad are PSO-SVM model,SVM model,variable fuzzy set theory and TOPSIS model respectively.Although there are differences between the grades,the inconsistent results of Farms with two different grades accounted for only 10% of the total evaluation results.The resilience of water resources is at the level o f I and II,which accounted for 53.3% of the total number of farms respectively,and the grade IV of water resources resilience of farms in Jiangsanjiang Administration are only for Farm Nongjiang and Farm Honghe.And its spatial distribution of water resources resilience has shown a increasing trend from southwest to northeast.According to the discriminant degree theory and the standardized scatter plot,the PSO-SVM model has the best reliability.(3)Based on the optimal measure model — PSO-SVM of agricultural water resources resilience,the resilience index of 15 farms in Jiansanjiang Administration of different years is calculated,and the resilience of each farm is graded.Spatial distribution maps of resilience levels in different years are drawn by using Arc GIS software,the temporal and spatial variation characteristics of resilience levels in different years are analyzed.According to the classification of each indicator,the relationship between the resilience level of each farm and the level of each indicator is obtained.And determine the resilience index grade of each farm is strong or weak in this area.The variability theory and grid unit analysis theory are used to analyze the interannual and water level changes of agricultural water resources resilience.The main obstacle indicators that restrict the recovery ability of agricultural water resources were diagnosed by obstacle index method.It can be seen that with the change of time,the spatial distribution of water resources resilience grade of farms in this region has changed significantly.The agricultural water resources resilience of farms in grade I and II accounts for more than half of the total number of farms in each year.Interannual changes are relatively stable,only the Qianfeng Farm and Shengli Farm are in strong fluctuations.According to the relationship between resilience level and farm level,the phenomenon that the index level of each farm is lower than the corresponding farm level,which requires special attention.(4)Construct a multi-objective cuckoo optimization algorithm to analyze the crop planting structure of the Jiansanjiang Administration,the main grain crops planted in this area in the past ten years are sequenced and analyzed.Finally,the rice,soybean and maize planted areas of each farm are selected as variables,and the total yield of crops,the maximum net output value of crops,the maximum ecological benefit and the maximum benefit of water production are selected as objective functions.The crop planting area,net crop output value per unit area,gross irrigation quota and resilience coefficient are selected as constraints.Through model optimization,it can be concluded that compared with the current year of 2017,the optimized economic benefits,ecological benefits and water production benefits of rice,soybean and corn planting areas in 2020 and 2025 are greatly improved.In addition,the agricultural planting structure is significantly constrained by the water resources resilience coefficient compared wi th the constraint of no restoring force coefficient,and the development of agricultural production in the region needs to be properly considered. |
PDF Full Text Request