滇池水质时间序列变化趋势识别及特征分析  

Identification of water quality trends in Lake Dianchi using an integrated STL and RSI approach

作　　者：梁中耀[1] 刘永[1] 盛虎[1] 杨永辉[1] 郭怀成[1] 赵磊 贺彬 

机构地区：[1]北京大学环境科学与工程学院水沙科学教育部重点实验室,北京100871 [2]云南高原湖泊流域污染过程与管理重点实验室,昆明650034

出　　处：《环境科学学报》2014年第3期754-762,共9页Acta Scientiae Circumstantiae

基　　金：国家水体污染控制与治理科技重大专项(No.2013ZX07102-006);国家自然科学基金(No.41222002)~~

摘　　要：应用STL(Seasonal-Trend Decomposition using LOESS)方法对1998—2010年滇池外海8个监测站点、10个水质指标的时间序列数据进行递归迭代,分解出低频率的趋势项、高频率的周期项及随机残差项.在此基础上计算稳态变化指数(RSI),对水质趋势的时、空变化特征进行识别和判定.结果表明,在分析的13年间,滇池外海8个监测站点各指标变化趋势总体一致;BOD5总体呈下降趋势,高锰酸盐指数(COD Mn)总体趋势显著上升,说明有机物可生化性下降;富营养化状态加剧,总氮(TN)总体趋势上升,并逐渐成为藻类的限制性营养盐,总磷(TP)浓度为劣V类,氨氮在氮元素形态中的主导地位正逐渐降低.研究结果表明,"十二五"期间滇池应重点控制氮磷、兼顾有机物污染.An integrated Seasonal-Trend Decomposition using LOESS （STL） and Regime Shift Index （RSI） approach was employed to identify the water quality trends in Lake Dianchi. The data of 10 water quality variables during 1998—2010 at eight monitoring stations in Lake Dianchi were analyzed. The trend components were filtered using the STL, and the characteristics of each trend component were analyzed using the RSI method. The model results showed that （a） the water quality at all monitoring stations demonstrated similar trends during 1998—2010; （b） Biochemical Oxygen Demand （BOD5） showed a decreasing trend while Permanganate Index （CODMn） concentrations showed a significantly increasing trend, suggesting that the organic matter in Lake Dianchi became less bio-degradable during the 13-year period; and （c） the eutrophication in Lake Dianchi intensified with time, with total nitrogen （TN） concentrations increasing through the 13-year period and nitrogen as the limiting nutrient gradually. The analysis also suggested that nitrogen and phosphorus reduction should be the focus for Lake Dianchi eutrophication control during the 12th Five-Year plan period.

关 键 词：时间序列 水质 趋势分解 滇池 稳态变化指数 

分 类 号：X171[环境科学与工程—环境科学]