College of Arts and Sciences, Northeast Agricultural University, Changjiang Street No. 600, Harbin, 150030, People's Republic of China.
Environ Monit Assess. 2024 Apr 25;196(5):477. doi: 10.1007/s10661-024-12625-y.
Heilongjiang reclamation area serves as a crucial hub for commodity grain production and strategic reserves in China, playing a vital role in maintaining national food security. Investigating the assessment of agricultural drought risk in this region can yield valuable insights into spatial and temporal variations in drought risk. Such insights can aid in formulating effective strategies for disaster prevention and mitigation, thereby minimizing food losses caused by drought disasters. This study employs a comprehensive indicator system comprising 17 indicators categorized into hazard, exposure, vulnerability, and resistance capacity. The projection pursuit model is applied to evaluate regional drought risk, while the PSO algorithm, optimized by the SSA algorithm, addresses the limitations of low local search ability and search accuracy during the large-scale search process of the PSO optimization algorithm. This study examines and compares the optimization and convergence capabilities of three algorithms: real number encoding-based genetic algorithm (RAGA), particle swarm optimization algorithm (PSO), and sparrow algorithm-based improved particle swarm optimization algorithm (SSAPSO). The analysis demonstrates that SSAPSO exhibits superior optimization performance and convergence properties, establishing it as a highly effective algorithm for optimization tasks. The findings reveal the following trends: over time, agricultural drought risk in Heilongjiang reclamation area has generally declined, with fluctuations observed in hazard and vulnerability, an increase in exposure, and a continuous enhancement of resistance capacity. Spatially, the western region exhibits significantly higher agricultural drought risk compared to the eastern region, primarily due to elevated hazard and vulnerability, coupled with lower resistance capacity. As the agricultural economy grows and agricultural expertise accumulates, the risk of agricultural drought decreases. However, variations in economic growth among different regions lead to diverse spatial distributions of risk.
黑龙江垦区作为中国商品粮生产和战略储备的重要基地，对维护国家粮食安全具有重要作用。研究该区域农业干旱风险评估，可以深入了解干旱风险的时空变化。这有助于制定有效的防灾减灾策略，最大限度地减少干旱灾害造成的粮食损失。本研究采用了一个包含 17 个指标的综合指标体系，这些指标分为危害、暴露、脆弱性和抵抗能力四个方面。采用投影寻踪模型对区域干旱风险进行评估，利用麻雀搜索算法优化粒子群算法（SSA-PSO），解决了粒子群优化算法（PSO）在大规模搜索过程中局部搜索能力和搜索精度低的问题。本研究对实数编码遗传算法（RAGA）、粒子群算法（PSO）和基于麻雀搜索算法的改进粒子群算法（SSAPSO）三种算法的优化和收敛能力进行了检验和比较。结果表明，SSAPSO 算法具有更好的优化性能和收敛特性，是一种有效的优化算法。研究结果表明：随着时间的推移，黑龙江垦区农业干旱风险总体呈下降趋势，危害和脆弱性呈波动变化，暴露度增加，抵抗能力不断增强。空间上，西部地区农业干旱风险明显高于东部地区，主要是由于危害和脆弱性较高，抵抗能力较低。随着农业经济的增长和农业专业知识的积累，农业干旱风险呈下降趋势。但是，不同地区经济增长的差异导致了风险的空间分布不同。
