长江水质的评价和预测

The Estimation and Forecast of the Yangtze River's Water Quality

本文利用长江流域近两年多主要城市水质检测数据,通过对原始数据进行归一化综合处理,确定了水质新的综合评判指标函数Ψ。在对整个长江流域所有观测站的位置关系作一定的简化假设后,得到长江综合评定函数值Ψ=0.4331,水质为良好,主要污染物为氨氮。通过建立污染浓度的反应扩散方程,本文用三种方法反演出未知的污染源强迫函数f(z,t),并对f(z,t)的三种数据加以综合分析,分别给出了高锰酸钾盐和氨氮污染源的主要分布地区。为了对长江未来水质污染发展趋势进行预测,本文建立了回归分析模型并对回归系数进行了F检验,结果是如果不采取有效的治理措施,长江可饮用水将逐年下降,且10年后可饮用水所占长江水总量的比例将不到50%。根据这一预测结果,我们进而使用二元线性回归模型,通过对各种不可饮用水进行综合考虑,得到如下结果:要在未来10年内使长江干流的不可饮用水(Ⅳ类和Ⅴ类水)的比例控制在20%以内,且没有劣Ⅴ类水,那么每年污水处理量至少为75.195亿吨。

In this paper we integrate the original data and get a new estimate function f（x, t）. After predigesting thc positions of all the observation stations along the Yangtze River, we obtain the estimate function f（x, t）, and find that the water quality is good, and the mostly contamination is ammonia and nitrogen. By establishing a reactive diffusion equation on the concentration of the pollute, we deduce the function of polluting source in three ways, then analyse its value and find the areas where potassium permanganate and ammonia and nitrogen distribute. For forecasting the tendency of the water quality, wc establish a regress model and do F-test on the coefficients in the regress model, then conclude that if we don＇t adopt any methods to control comtaminations, the quantity of drinking water would decline year by year, and the gross quantity of Yangtze River＇s drinking water will be half of it ten years later. So we can use bi-element linear regress model to consider all kinds of water which is not able to drink and make following conclusion： in ten years if we make the proportion of nondrinking water to be less than 20, then the quantity of sewage must be treated with more than 75.195 hundred million tons per year.