A measured bias correction model for the estimation of river contaminant loads from discharge data

Estimation of contaminant loads carried by rivers is important for water quality management. A study was carried out for estimating mean monthly solids load from river discharge data by application of the rating curve equation and method of moments. The results show both the untransformed and log transformed discharge data underestimate the solids load with relative standard deviation of residuals of 60% for untransformed discharge data and 55.6% for log-transformed discharge data. Application of measured load bias correction modelling reduced the relative standard deviation of residuals to 10.8% for the untransformed discharge data and 9% for the log transformed discharge data. The measured solids load bias was reduced from 1.6 to around 1.0 on average. The residuals may be reduced further by using mean discharge derived from data through optimisation. The method has limitation where there is weak relationship between contaminant load and river discharge where further research is needed.