Li Wei-Feng, Mao Jing-Qiao
Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences, Beijing 100085, China.
Huan Jing Ke Xue. 2011 Nov;32(11):3200-6.
An integrated eutrophication assessment framework is developed for lakes and reservoirs based on an ecogeographical classification method and an artificial neural network model. Using the USEPA Nutrient Criteria Database as the basic reference and considering the ecogeographical characteristics of Chinese lakes and reservoirs, a simple eutrophication assessment criterion considering the ecogeographical characteristics is proposed for the first time. This criterion places the emphasis on the determination of critical values of key parameters for various regions. Moreover, an artificial neural network (ANN) assessment model is developed, considering the complexity and nonlinearity of eutrophication process. It is found that this ANN assessment model offers the advantage to assess with more accuracy the trophic status in nitrogen-limited water bodies. Integrating such two assessment methods can establish a simple but general eutrophication assessment framework; verification with 30 lakes and reservoirs shows that it can be served as a reliable and cost-effective tool for aquatic environmental management.
基于生态地理分类方法和人工神经网络模型，开发了一种针对湖泊和水库的综合富营养化评估框架。以美国环保署营养标准数据库为基本参考，并考虑中国湖泊和水库的生态地理特征，首次提出了一种考虑生态地理特征的简单富营养化评估标准。该标准着重于确定各区域关键参数的临界值。此外，考虑到富营养化过程的复杂性和非线性，开发了一种人工神经网络（ANN）评估模型。结果发现，该ANN评估模型具有更准确评估氮限制水体营养状态的优势。将这两种评估方法相结合，可以建立一个简单但通用的富营养化评估框架；对30个湖泊和水库的验证表明，它可作为水生环境管理的可靠且具有成本效益的工具。
