Algalbloom prediction models for Liuhai-lake in Beijing city
-
摘要: 采用决策树方法和非线性回归方法建立湖泊水华预警模型。决策树方法预测水华爆发时机,非线性回归方法预测水华爆发强度,并运用信号灯显示方法,划分出水华爆发的预警区间。以北京六海为例,模型结果表明来水水量Q,温度T和总磷浓度是影响“六海”湖泊水华爆发的主要影响因子,选择叶绿素a(Chl-a)<30 μg/L的预警信号为绿色,30 μg/L
60 μg/L为红色。当每月来水量Q>79.0万m3或来水量Q<79.0万m3,水温<13.4℃,预警指标为绿色;Q<79.0万m3,水温T>13.4℃,水华预警为黄色;Q<38.7万m3时,T>23.25℃,TP>0.13 mg/L,水华预警为红色。对模型结果分类进行了验证。结果表明:模型对于限制因素发生变化时的水华预测结果更为准确,并且结构简单,输入输出关系明显,结果易于解释。 Abstract: The algalbloom prediction models are constructed by using the decision trees to qualitatively predict bloom timing and use the nonlinear piecewise regression to quantitatively predict bloom intensity.The traffic light systems are used as the indicator for degree of algal bloom.Liuhai-lake in Beijing city is used an example.The constructed model indicates that the water inflow,the temperature and total phosphorus are the most impact factors on the algalbloom in Liuhai-lake.The concentration of Chl-a<30 μg/L is labeled as green,30 μg/L60 μg/L as red.When water inflow Q>79.0×104m3 or Q<79.0×104m3 and water temperature<13.4℃,the indicator is green;when Q<79.0×104m3, water temperature T>13.4℃,the indicator is yellow;and when Q<38.7×104m3,water temperature>23.25℃,TP>0.13 μg/L,the indicator is red.The model is test by an independent dataset from the same area,the predicted blooming time error rate and the error of predicted bloom intensity are presented in the paper.The model has great advantages to deal with the common problem in algal-blooms.It's more accurate when the limiting factor is changing.And the structure is understandable and easy to interpret. -
Key words:
- algalbloom /
- prediction /
- decision trees /
- nonlinear regression /
- traffic light system /
- Liuhai-lake in Beijing city
-
[1] 卢小燕,徐福留,詹巍,等.湖泊富营养化模型的研究现状与发展趋势[J].水科学进展,2003,14(6):792-798. [2] 夏军,窦明.水体富营养化综合水质模型及其应用研究[J].上海环境科学,2000,19(7):302-308. [3] 彭虹,郭生练.汉江下游河段水质生态模型及数值模拟[J].长江流域资源与环境,2002,11(4):363-369. [4] 金腊华,徐峰俊.水环境数值模拟与可视化技术[M].北京:化学工业出版社,2004. [5] Walter M.Predicting eutrophication effects in the Burrinjuck reservoir(Australia) by means of the deterministic model SALMO and the recurrent neural network model ANNA[J].Ecological modeling,2001,146:97-113. [6] 周勇,刘凡,吴丹,等.湖泊水环境预测的原理和方法[J].长江流域资源与环境,1999,8(3):305-311. [7] Maier H R,Dandy G C,Burch M D.Use of artificial neural networks for modeling cyanobacteria Anabaena spp.In the river murray,south Australia[J].Ecological modeling,1998,105:257-272. [8] Gurbuz H.Predicting dominant phytoplankton quantities in a reservoir by using neural networks[J].Hhydrobiologia,2003,504:133-141. [9] 裴洪平,罗妮娜,蒋勇.利用BP神经网络方法预测西湖叶绿素a的浓度[J].生态学报,2004,24 (2):246-251. [10] Solomatine D P,Dulal K N.Model tree as an alternative to neural network in rainfall-runoff modeling[J].Hydrological Science Journal,2003,48:399-411. [11] Chen Q,Mynett A E.Predicting phaeocystis globosa bloom in Dutch coastal waters by decision trees and non-linear piecewise regression[J].Ecological Modeling,2004,176:277-290. [12] Mingers J.An Empirical comparison of selection measures for decision-tree induction[J].Machine learning,1989,3:319-342. [13] Quinlan J R.C4.5:Programs for machine learning[M].San Mateo,California:Morgan Kaufrmann,1993. [14] 张文彤.SPSS11统计分析教程高级篇[M].北京:希望电子出版社,2002. [15] 韩秀娣.非点源污染最佳管理实践[J].上海环境科学,2000,19(3):102-104. [16] Webb A A,Erskine W D.A practical scientific approach to riparian vegetation rehabilitation in Australia[J].Journal of environmental management,2003,68:329-341. [17] 全为民,沈新强,严力蛟.富营养化水体生物净化效应的研究进展[J].应用生态学报,2003,14(11):2 057-2 061. -

计量
- 文章访问数: 98
- HTML全文浏览量: 17
- PDF下载量: 465
- 被引次数: 0