基于贝叶斯公式的湖泊富营养化随机评价方法及其验证

为了准确评价湖泊所处的营养状态，进而为湖泊富营养化的防治提供科学依据，提出了基于贝叶斯公式的湖泊富营养化随机评价方法。该法通过计算单个水质指标属于某个评价级别的概率，由最大似然分类原则确定单个水质指标的评价级别，并采用加权平均法推求湖泊富营养化多个水质指标的综合评价级别，具有方法简单、计算工作量小的特点。经全国30个湖泊实测水质资料的验证，随机评价方法与模糊评价方法和灰色评价方法相比较，有29个湖泊评价结果的差异小于等于一个级别，其比例为96.6％，说明这三种富营养化评价方法的计算精度相当。根据各项评价指标与叶绿素的相关系数，计算各项水质指标在富营养化评价中所占权重的方法，适用于浮游植物型湖泊，如果采取等权重方法，则适用于任何类型的湖泊。确定湖泊富营养化评价的水质指标个数和种类的选择具有不确定性，确定湖泊富营养化的评价级别和各个级别对应的水质标准浓度也具有不确定性，因此湖泊富营养化的评价方法具有不确定性。应采用多种途径对湖泊富营养化的评价方法进行研究，并取长补短，使评价结果更趋合理。

A LAKE EUTROPHICATION STOCHASTIC ASSESSMENT METHOD BY USING BAYESIAN FORMULA AND ITS VERIFICATION

In order to correctly evaluate the trophic state of a lake and then to establish scientific basis for the prevention of lake eutrophication, a lake eutrophication stochastic assessment method (LESAM) by u-sing Bayesian formula is proposed. In this method, according to the maximum likelihood ordering principle, the assessment order of each water quality index of an actual lake was firstly derived through calculating the probability of each of those water quality indices belonging to the assessment order used in the lake eutrophication assessment criteria. Secondly, the assessment order of lake eutrophication is derived by u-sing the weighing average method. The characteristics of LESAM are that the method is simple, and the computation work is moderate. LESAM, together with other two methods, i. e. the lake eutrophication fuzzy assessment method (LEFAM) and the lake eutrophication gray assessment method (LEGAM), has been tested against the measured data of 30 lakes all over the country. It is found that the difference of assessment results among those three methods is not greater than 1 order in 29 lakes, i. e. 96. 6 % of the total lakes. It is concluded that all these three methods (LESAM, LEFAM, and LEGAM) have comparable computation precision. Though in lake eutrophication assessment we can calculate the weight of each water quality index according to the correlation coefficient between each water quality index and chlorophyll, this method only suits the lakes of phytoplankton type. As to the equally weighting average method, it can be used for any kind of lakes. For there is a lot of uncertainty involved both in determining the number and type of water quality index in lake eutrophication assessment and in determining the assessment order and its corresponding standard water quality concentration, the lake eutrophication assessment methods also have uncertainty. It is suggested that the lake eutrophication assessment should be studied with various approaches in order to overcome the weakness of one particular method and to make assessment results more reasonable.