Akademik Veri Yönetim Sistemi
J. Int. Environmental Application & Science, cilt.19, sa.2, ss.57-68, 2024 (TRDizin)
Water quality control, especially in large-scale monitoring regions or networks, requires easy and automatic processes for detecting potential outliers in a reproducible manner. This study focuses on removing outlier values from a dataset collected by an online monitoring station on the Yeşilırmak River between 2007 and 2009. Seven different parameters were evaluated: dissolved oxygen (luminescence dissolved oxygen, LDO), temperature, pH, conductivity, total organic carbon (TOC), nitrate nitrogen (NO3-N), and ammonium nitrogen (NH4-N). Five methods – median, mean, Grubbs’, generalized extreme studentized deviate (GESD), and interquartile range (IQR) – were used for outlier removal. The developed models were integrated into a graphical user interface (GUI) in the MATLAB environment, facilitating practical and easy access. This study enables users to input any dataset into the software and remove outlier values using various methods in a few steps, thus preparing the data for modeling studies. It was observed that the median algorithm removed the most data points among the outlier data-removal methods.