Abstract:
Missing rainfall data is a common problem in the practice of hydrology. To complete the missing rainfall data, there are several methods that can be used. Application of a method in a particular location is not guaranteed to produce a satisfactory outcome. For this study, the effectiveness of a model at a specific location is necessary. This study examined the effectiveness and behavior for Standardize Nonlinear Multiple Correlation (MNSC) methods in the implementation of the Watershed in Indragiri and Rokan. For this study, 10 years monthly rainfall data was used. Calculation process was done by making several variations of the simulation analysis of data relating to the condition of filling stations. The results of the calculation will be evaluated by using two categories of errors which are absolute errors and the comparison of statistical parameters of the count data from the original data. The results showed the error evaluation value was greater than that required by the method MNSC less than 25%. Behavior of the model indicates that the data length and sequence of the data does not affect the size of the errors that are generated, and there was no guarantee that a high correlation coefficient between the variable filling station and filled a small error value.
