Journal of Engineering and Applied Sciences

Year: 2019
Volume: 14
Issue: 9 SI
Page No. 10716 - 10722

Evaluation of Filling Missing Data Technique in Hydraulic Series

Authors : Adnan K. Shathir

Abstract: The missing data is one of the main problems in the hydraulic research, specially when need large amounts of data, the missing data is made the series of data discrete and not sufficient to research. This problem is very serious in Iraq because of large amount of the discrete time series due wars and other reasons. So, in order to solve this problem in Iraq the simple technique of filling missed data is chosen with the assumption that no information about the relations between the nearest stations and the target station. So, the target stations with missing data is the only series used to estimate missing data. The reliability of the values of estimated data were evaluated by verification of 30 gaging station inflow discharges for United State rivers with different locations and different hydraulic properties. In this study, the year 1975 is chosen randomly to be considered as a missing data. So, the data of this year was removed from the monthly discharge time series for each river and use three approaches of filling missed data, the first method use the mean of the month of time series as estimated value for missing month, the second method use the interpolation between two months to estimate the missing data while the third method use the nearest 2 months to modify the value of mean of missing month. Then we compare the computed values with the actual data to know how these values have percentage of error to decide how it’s proper or not to use in analysis research. From the results, the mean method and modified mean method for monthly time series can be used with confidence but it depends on what reason these estimated data will be used in research analysis.

How to cite this article:

Adnan K. Shathir , 2019. Evaluation of Filling Missing Data Technique in Hydraulic Series. Journal of Engineering and Applied Sciences, 14: 10716-10722.

Design and power by Medwell Web Development Team. © Medwell Publishing 2024 All Rights Reserved