期刊文献+

基于时间序列模型与BP神经网络的深圳近岸海域富营养化预测 被引量:7

Prediction of Eutrophication in Shenzhen Coastal Waters Based on Time Series Model and BP Neural Network
下载PDF
导出
摘要 根据2013~2017年夏季(8月)深圳近岸海域水质监测数据资料,采用营养质量指数法对深圳近岸海域富营养化状况及其变化趋势进行评价,并建立时间序列预测模型与BP人工神经网络模型分别对该海域2018年无机氮、活性磷酸盐、化学需氧量和叶绿素a含量进行预测,并根据预测结果利用营养质量指数法对2018年深圳海域富营养化水平进行预测,同时根据预测结果对深圳海域营养盐结构进行分析,识别东西部海域的富营养化限制因子,据此提出富营养化水平的改善建议。研究结果表明:2013~2017年深圳近岸海域富营养化水平一致表现为西部海域偏高,东部海域偏低的特征,其中珠江口海域在5年间的平均富营养化水平最高,其余依次为深圳湾>大鹏湾>大亚湾。2018年富营养化水体基本出现在西部海域,东部海域整体富营养化水平较低,以贫营养水平为主。富营养化高风险区主要分布在西部海域,其中茅洲河口与深圳湾海域富营养化风险尤为高,建议在上述区域进行富营养主要因子无机氮和活性磷酸盐的管控,以期降低富营养化发生概率。针对东部海域沙头角湾和坝光海域表现为中营养水平的情况,建议对上述海域进行无机氮和活性磷酸盐的防治,以达到保持海域营养水平现状的目的。 Based on the water quality monitoring data in the summer of Shenzhen coastal water from 2013 to 2017, the eutrophication level was evaluated by the method of nutrition quality index and its change trend in coastal waters of Shenzhen was analyzed. At last, the content of inorganic nitrogen, active phosphate, chemical oxygen demand and chlorophyll-a of Shenzhen nearshore area was predicted by time series forecasting model and artificial neural network, then the eutrophication level of 2018 was assessed. In the mean while, the nutrient structure of Shenzhen nearshore area was analyzed to recognize the limiting factor of eutrophication. The results show that, the level of eutrophication in the coastal waters of Shenzhen was unanimously high in west sea area and low in eastern sea area from 2013 to 2017, and the average eutrophication level reached the highest in the Pearl River estuary in 5 years, and others successively were Shenzhen bay, Dapeng bay and Daya bay. Eutrophication water basically appeared in the western waters, and the eutrophication level in the eastern waters was relatively low in 2018.The high risk areas of eutrophication are mainly distributed in the western waters, among which, the risk of eutrophication is particularly high in Maozhou estuary and Shenzhen Bay, and suggested that the inorganic nitrogen and active phosphate should be controlled in Maozhou estuary and Shenzhen Bay to reduce the probability of eutrophication. In order to maintain the status of nutrient level in this sea area, the policy of putting prevention should carried out in Shatoujiao bay and Baguang sea area in the eastern sea.
作者 陈芸 周连宁 唐俊逸 赵振业 CHEN Yun;ZHOU Lian-ning;TANG Jun-yi;ZHAO Zhen-ye(IER Environment Protection Engineering Technique Co., Ltd.,Shenzhen 518057, China;Shenzhen Key Laboratory forCoastal andAtmospheric Research, Shenzhen 518057, China)
出处 《科学技术与工程》 北大核心 2019年第4期66-73,共8页 Science Technology and Engineering
基金 深圳市人居环境委员会科研计划资助
关键词 时间序列模型 BP神经网络 深圳海域 富营养化 time series model back-propagation neural network Shenzhen sea area eutrophication
  • 相关文献

参考文献18

二级参考文献244

共引文献724

同被引文献80

引证文献7

二级引证文献27

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部