New method for diagnosing resilience of agricultural soil-water resource composite system: Projection pursuit model modified by sparrow search algorithm

• Published in: Journal of hydrology (Amsterdam) Vol. 610; p. 127814
• Main Authors: Xu, Dan, Liu, Deping, Liu, Dong, Fu, Qiang, Huang, Yan, Li, Mo, Li, Tianxiao
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 01.07.2022
• Subjects: Agricultural soil–water resource composite system; algorithms; China; Driving mechanism; entropy; evolution; gross domestic product; income; Logarithmic mean Divisia index; paddies; Passeriformes; Projection pursuit; Resilience; simulation models; soil water; Sparrow search algorithm; sustainable development; China; Agricultural soil–water resource composite system; Logarithmic mean Divisia index; Sparrow search algorithm; Projection pursuit; Driving mechanism; Resilience
• ISSN: 0022-1694; 1879-2707
• DOI: 10.1016/j.jhydrol.2022.127814

•Constructing an analysis model of resilience driving mechanism based on SSA-PP-LMDI.•Identify key driving factors and determine the driving effects of key driving factors.•Prediction of the future evolution trend of resilience based on key driving factors. Considering the driving mechanism of agricultural soil–water resource composite system (ASWRS) resilience, the Jiansanjiang Branch of China Beidahuang Agricultural Reclamation Group Co., Ltd., is selected as the research object. A projection pursuit (PP) model improved by the sparrow search algorithm (SSA) was proposed to measure the key drivers of resilience, and the driving mechanism of resilience was analysed in combination with the logarithmic mean Divisia index (LMDI). The performance of the built model and the rationality of the results were verified, thus providing a new horizon for the healthy and sustainable development of agricultural water and soil resources and the environment. The results indicated that the SSA-PP model constructed in this paper was superior to PP models based on genetic algorithm (GA) optimization (GA-PP) and particle swarm optimization (PSO-PP) in regard to the solution speed and optimization ability, and these three models yielded similar conclusions. Compared with those of the entropy weight (EW) method, the driving results of the SSA-PP model are more significant, and the screened key driving factors are more reasonable. It was finally determined that the impact on the resilience of the ASWRS was the result of the joint action of economic and agricultural systems. The key driving factors included the per capita gross domestic product (GDP), paddy field proportion, per capita net income and grain output per unit area. According to the LMDI results, the economic level, grain output and paddy field proportion increased, which played a positive role in driving resilience. With the use of the autoregressive integrated moving average (ARIMA) model to simulate the future evolution of the key driving factors, it was predicted that the resilience level will increase year by year in the future. It was found that accelerating the economic development process, increasing the grain output, implementing rational planning and adjusting the agricultural planting structure are effective ways to enhance resilience.