三峡水库次级河流高频自动监测与数据挖掘
2019-10-21刘伟吴庆梅邓力杨兵
刘伟 吴庆梅 邓力 杨兵



摘要 2013年对三峡水库次级河流10个点位7个参数进行了原位高频自动监测研究。结果显示浊度的季节性差异不显著(P>0.05),为光合生物提供了良好的光照基础条件。水温、电导率、溶解氧、pH表现出季节性极显著差异(P<0.001),叶绿素a、蓝绿藻的季节性差异显著(P<0.01)。pH与蓝绿藻显著正相关(P<0.05),溶解氧与叶绿素a、蓝绿藻显著正相关(P<0.01)。pH的变化幅度、叶绿素a和蓝绿藻的最大值,单独或结合分析都可预警水环境发生重大量变,但是否发生质变,则需要进一步研究算法。
关键词 三峡水库;次级河流;高频自动监测;数据挖掘
中图分类号:X524 文献标识码:A 文章编号:2095-3305(2019)05-057-03
DOI: 10.19383/j.cnki.nyzhyj.2019.05.023
High Frequency Automatic Monitoring and Data Mining of Secondary Rivers in the Three Gorges Reservoir
LIU Wei et al(Ecological and Environment Monitoring Center of Chongqing,Chongqing 401147)
Abstract In 2013,the in-situ high-frequency automatic monitoring of 7 parameters at 10 points in the secondary river of the Three Gorges Reservoir was carried out. The results showed that the seasonal difference of turbidity was not significant (P>0.05),which provided a good light base for photosynthetic organisms. The seasonal differences of water temperature,conductivity,dissolved oxygen and pH value were significant (P<0.001),the seasonal differences of chlorophyll a and blue-green algae were significant (P<0.01),and the pH value was positively correlated with that of blue-green algae (P<0.05). Dissolved oxygen was positively correlated with chlorophyll a and blue-green algae (P<0.01). The mean value of dissolved oxygen,the variation range of pH,the abnormal value of chlorophyll a,the maximum value of blue-green algae,alone or in combination with the analysis,could be used to predict the significant quantitative change of water environment,but whether the qualitative change occurs or not,the algorithm needs to be further studied.
Key words Three Gorges Reservoir;Secondary river;High frequency automatic monitoring;Data mining
三峽水利工程蓄水后,许多学者对次级河流水体富营养化程度加重[1]这一现象进行了研究。有以单一季节水华藻类优势种、毒性及营养限值为研究对象[2-3],以单次水华事件及其发生时的理化条件为研究对象[4],以特定水期氮、磷分布特征为研究对象[5],也有以每月定时定点河流的生物和理化因子变化为研究对象[6]。均采取现场采样、实验室分析的传统研究手段。从时间序列看,这种单次观测结果数据“离散”,难以捕捉水华的暴发-消亡生态过程。近年来,有学者在太湖研究了原位高频自动监测技术[7],并把它作为水华预警监测系统的子系统加以应用,但仍需要完善且保持稳定运行[8]。笔者将该技术全天候应用到三峡库区次级河流中,为水华研究获取了长时间序列基线数据,为预警算法的研究打下基础。
1 材料与方法
2013年在三峡水库次级河流回水区设置10个监测断面,分别为S1金竹滩、S2红旗、S3老龙潭、S4双江、S5清水码头、S6草堂、S7大昌、S8白水口、S9龙头山、S10葡萄坝,点位设置和水位变化见图1。五参数(水温、pH、电导率、浊度、溶解氧)直接反映水质,连续监测成本低且与其他指标存在一定相关性,因而在预警监测中不可或缺[9]。荧光……
