河流污染物通量估算方法筛选及误差分析

准确估算污染物通量对河流污染物总量控制和水质保护具有重要意义,但常规监测方法通常不能进行连续采样,数据相对比较稀缺,因此,选择恰当的采样间隔和通量估算方法,并对方法可能带来的误差进行估算,对提高通量估算的可靠性具有十分重要的意义.基于此,以江西赣江滁槎水质自动站2005—2007年3个水文年的水量和水质资料为基础,采用Monte Carlo方法模拟时间间隔分别为2、3、5、6、10、15和30 d的河流水质离散采样方案,并计算每种采样方案下的通量.同时,采用偏差(系统误差)和不精确度(离散程度)两个指标,比较了A、B、C、D、E 5种常规通量计算方法的误差分布,并建立了各算法误差随时间间隔变化的相关性曲线,以对河流污染物通量估算方法进行筛选.研究表明,滁槎断面COD_Mn采用瞬时值C_i与时段平均流量 Q_p乘积的方法计算年通量更准确;而NH₄⁺-N由于瞬时通量与流量相关性不显著,采用时段瞬时通量平均计算年通量更准确.

Study on the selection and error analysis of riverine pollutant flux estimation methods

Accurate estimation of riverine pollutant flux is very important for river total load control and water quality protection.Routine water quality monitoring can not be made on line in a lot of occasions and the data scarcity is usually an inevitable problem.Therefore,selection of proper sampling intervals and evaluation of the possible errors of different flux calculation methods are essential in improving the flux estimation reliability.Based on the 2005—2007 hydrological and water quality data of Ganjing Chucha online monitoring station in Jiangxi Province,the pollutant flux was calculated by the intervals of 2,3,5,6,10,15 and 30 d with Monte Carlo methods to simulate different sampling interval schemes.The error distribution characteristics of the 5 methods as A,B,C,D,E were compared with two indices,i.e.bias(or systematic error) and imprecision(degree of dispersion),then the error versus sampling interval curves were drawn and applied to evaluate the performance of different flux estimation methods.It showed that in Chucha section it was more accurate to estimate flux of CODMn with instantaneous Ci times temporal average flow,but the method of instantaneous flux average was better than others for NH+4-N because of the poor relationship between its flux and the flow.