Temporal Classification to Improve the Precision of Suspended Sediment Rating Curve (Case Study: Gorganrood Basin)

In most cases, the suspended sediment rating curve model is not very accurate for estimating the suspended sediment load of the river. Various studies have shown that the relationship between river discharge and suspended sediment has temporal changes. In this study, three seasonal, monthly, and wet–dry-flood season models were developed based on the concept of temporal variability to improve the precision of the suspended sediment rating curve model. To evaluate the developed models, data from 30 hydrometric stations in the Gorganrood basin area in northern Iran were used and the index of agreement criterion was used to compare the precision of these models. The geometric means of the index of agreement and coefficient of determination in the basic model are equal to (0.64 and 0.59) and in the developed models are equal to (0.75 and 0.59), (0.79 and 0.62), and (0.76 and 0.70), respectively. These values indicate that all three developed models significantly improved the precision of the base model. Also, the coefficient of variation of the index of agreement in the basic model is equal to 0.15, while this value has decreased to 0.09 in all three developed models, which indicates the better consistency of the developed models in comparison to the base model. On the one hand, the difference in the precision of the results in the developed models is limited. On the other hand, the monthly model is not highly reliable due to various parameters. Therefore, it seems that the wet–dry-flood season model can be considered a suitable model to improve the base model of the suspended sediment rating curve of the river.