Assessment of water quality classes using self-organizing map and fuzzy C-means clustering methods in Ergene River, Turk
- PDF / 4,941,950 Bytes
- 10 Pages / 547.044 x 736.903 pts Page_size
- 46 Downloads / 170 Views
(2020) 192:638
Assessment of water quality classes using self-organizing map and fuzzy C-means clustering methods in Ergene River, Turkey Ersin Orak · Atilla Akkoyunlu · Zehra Semra Can
Received: 28 February 2020 / Accepted: 20 August 2020 © Springer Nature Switzerland AG 2020
Abstract Surface water is one of the primary sources for drinking, irrigation, and industrial activities in Ergene River, Turkey. However, its quality has deteriorated due to the point and non-point pollution sources. Therefore, an appropriate assessment of surface water quality is very important. Water quality classification is calculated separately for each quality parameter in Turkey. An overall assessment of surface water quality is essential for water management. In this study, selforganizing maps (SOMs) and fuzzy C-means clustering (FCM) methods have been used for assessing surface water quality in the Ergene River. Seven water quality parameters have been considered as important indicators to evaluate water quality status in 7 observation points located in the river, covering the period from 1985 to 2013. Keywords Water quality assessment · Self-organizing map · Fuzzy C-mean clustering · Ergene basin
E. Orak () · Z. S. Can Marmara University, Istanbul, Turkey e-mail: [email protected] A. Akkoyunlu Bo´gazic¸i University, Istanbul, Turkey
Introduction The importance of national and global water resources has increased due to decreasing available water potential and rapidly increasing world population (Evsahibio˘glu et al. 2010). Anthropogenic activities have deteriorated the water quality, assessment of which is particularly important for proper water management. Various methods have been applied for calculation of water quality, such as mathematical modelling, statistical techniques, and soft computing methods (Bilgin and Konanc¸ 2016). Prior research has shown that simulations based on deterministic water quality models often lead to inaccurate results, as these models can neither capture the complexity of water quality which depends on many factors nor deal with the imperfections and uncertainty in the underlying data. In the literature, multivariate statistical methods have been used frequently to calculate surface water quality (˙Is¸c¸en et al. 2009; Bilgin and Konanc¸ 2016). Conventional clustering methods which use Boolean logic also proved insufficient to process ecological data (Salski 2007). Recent years have seen the adoption of various soft computing methods to address the limitations of deterministic models: artificial neural networks (ANNs), adaptive neurofuzzy inference system (ANFIS), fuzzy C-means
638 Page 2 of 10
Environmental Monitoring and Assessment
clustering (FCM), support vector machines (SVM), and genetic programming (GP) are some of the most commonly used soft computing methods applied to water quality modelling (Najafzadeh et al. 2018). The self-organizing maps (SOMs) and fuzzy logic have been used inclusively for the assessment of water quality from the complex data (Juntunen et al. 2013; Voyslav
Data Loading...
