闫雪, 黄法融, 李倩, 周宏飞, 李兰海. 基于生态系统服务和PSO-SOFM神经网络的中亚水土热资源匹配分区[J]. 中国生态农业学报(中英文), 2021, 29(2): 241-255. DOI: 10.13930/j.cnki.cjea.200412
引用本文: 闫雪, 黄法融, 李倩, 周宏飞, 李兰海. 基于生态系统服务和PSO-SOFM神经网络的中亚水土热资源匹配分区[J]. 中国生态农业学报(中英文), 2021, 29(2): 241-255. DOI: 10.13930/j.cnki.cjea.200412
YAN Xue, HUANG Farong, LI Qian, ZHOU Hongfei, LI Lanhai. Regionalization of the matching degree of water, soil, and heat resources in Central Asia based on ecosystem services using PSO-SOFM neural network[J]. Chinese Journal of Eco-Agriculture, 2021, 29(2): 241-255. DOI: 10.13930/j.cnki.cjea.200412
Citation: YAN Xue, HUANG Farong, LI Qian, ZHOU Hongfei, LI Lanhai. Regionalization of the matching degree of water, soil, and heat resources in Central Asia based on ecosystem services using PSO-SOFM neural network[J]. Chinese Journal of Eco-Agriculture, 2021, 29(2): 241-255. DOI: 10.13930/j.cnki.cjea.200412

基于生态系统服务和PSO-SOFM神经网络的中亚水土热资源匹配分区

Regionalization of the matching degree of water, soil, and heat resources in Central Asia based on ecosystem services using PSO-SOFM neural network

  • 摘要: 水土热资源匹配度分区研究对于区域农业规划具有重要意义。中亚地区长期以来缺乏合理的水土热资源管理,已引发了一系列资源环境问题,严重威胁该地区农业生产。目前的研究也较少关注中亚水土热资源匹配分区模式。本研究利用遥感数据,通过量化4种主要生态系统服务(植被固碳、土壤保持、水源供给与涵养及生物多样性保护)的时空分布特征,结合PSO-SOFM(particle swarm optimization,PSO;self-organizing feature map,SOFM)神经网络模型实现中亚水土热资源匹配度分区,并利用Spearman秩相关分析探索不同匹配度分区与生态环境因子的关系,应用偏相关分析确定气温和降水量对中亚地区生态系统服务的影响。结果表明,中亚生态系统服务总体呈东南高、西北低的空间格局,沿山地-绿洲-荒漠方向递减。在2000-2015年间,各类生态系统服务均有不同程度变化,其中植被固碳和土壤保持呈显著下降的面积占整个中亚的84.81%和84.82%;水源供给与涵养以及生物多样性保护服务呈显著下降的面积较少,占比分别为69.48%和19.8%,且这两种生态系统服务在个别地区有增加趋势。PSO-SOFM神经网络模型在中亚水土热资源匹配度分区中表现良好,根据生态系统服务值空间模式,中亚水土热资源匹配度可被划为5大类21个子类分区。在空间尺度,各类匹配度分区之间生态系统服务值有显著差异,降水是影响生态系统服务和匹配度高低的重要限制因子,而气温和土壤因素影响较弱;在时间尺度,降水和各生态系统服务值间呈显著正相关关系的范围更广,而气温对生态系统服务值有显著影响的区域主要集中在哈萨克斯坦北部草地-半荒漠生态敏感区、中亚荒漠生态脆弱区、中亚中部半荒漠生态敏感区以及巴特赫兹-卡拉比尔半荒漠生态敏感区等地。而在其他区域,气温和降水量并非决定生态系统服务值高低的主要因素,生态系统服务值的变化可能与土地开发利用模式有关。结合不同匹配度分区的生态地理条件,本研究可为中亚地区水土资源开发利用、农牧业发展以及生态环境保护提供有用信息。

     

    Abstract: Regionalization of the matching degree of water, soil, and heat resources is of great significance for regional agricultural planning. The long-term unreasonable management of water, soil, and heat resources has caused regional resource shortages and environmental problems in Central Asia, which seriously threatens agricultural production in this region. However, few studies have investigated the regionalization patterns of the matching degree of water, soil, and heat resources in Central Asia. In this study, the spatio-temporal patterns of four ecosystem services, including vegetation carbon sequestration, soil conservation, water supply and conservation, and biodiversity conservation, were quantified by using remote sensing data. Combined with the Particle Swarm Optimization (PSO) and Self-Organizing Feature Map (SOFM) neural network, the regionalization of the matching degree of water, soil, and heat resources was examined. The relationships among various eco-environmental factors of different matching degree zones were assessed using Spearman's rank correlation analysis. The effects of temperature and precipitation on ecosystem services in Central Asia were analyzed by using partial correlation analysis. The results showed that the ecosystem services were generally high in the southeast while low in the northwest, decreasing from the mountains to the oases and the deserts. The four ecosystem services showed different degrees of change from 2000 to 2015 in Central Asia. Areas with significantly reduced vegetation carbon sequestration and soil conservation accounted for 84.81% and 84.82% of Central Asia, respectively, and areas with significantly reduced water supply and conservation and biodiversity conservation accounted for 69.48% and 19.8% of Central Asia, respectively. However, the ecosystem services from water supply and conservation and biodiversity conservation increased in some areas. The PSO-SOFM neural network model performed well in the regionalization of the matching degree of water, soil, and heat resources in Central Asia. The matching degree of water, soil, and heat resources in Central Asia can be divided into five categories with 21 sub-categories according to the patterns of ecosystem services. At the spatial scale, there were significant differences in the ecosystem services among different matching degree zones. Precipitation was the most important limiting factor affecting the ecosystem service values and matching degree, whereas the effects of temperature and soil properties were less important. At the temporal scale, the areas with a significant positive correlation between precipitation and ecosystem services were larger. The significant effect of temperature on ecosystem service values was mainly concentrated in ecological sensitive zone of northern Kazakh steppe and semi-desert, ecological fragile zone of desert in Central Asia, ecological sensitive zone of central semi-desert in Central Asia and ecological sensitive zone of semi-desert in Badghyz and Karabil. In other regions, temperature and precipitation were not the main factors affecting ecosystem services. Changes in the ecosystem service values may be related to land use types. Combined with the ecological and geographical conditions of different matching degree zones, this study provides useful information for the development and utilization of water and land resources, agriculture and animal husbandry development, and environmental protection in Central Asia.

     

/

返回文章
返回