Maximum stream temperature estimation of Degirmendere River using artificial neural network

KARACOR A. G., Sivri N., UÇAN O. N.

JOURNAL OF SCIENTIFIC & INDUSTRIAL RESEARCH, vol.66, no.5, pp.363-366, 2007 (SCI-Expanded) identifier identifier

  • Publication Type: Article / Article
  • Volume: 66 Issue: 5
  • Publication Date: 2007
  • Journal Name: JOURNAL OF SCIENTIFIC & INDUSTRIAL RESEARCH
  • Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus
  • Page Numbers: pp.363-366
  • Istanbul University Affiliated: Yes

Abstract

Stream temperature determines the rate of the decomposition of organic matter and the saturation concentration of dissolved oxygen. Combined with industrial waste, stream temperature becomes a crucial parameter. Therefore, estimation of maximum stream temperature is very important, especially during summertime when the high temperatures may become dangerous for the habitat of rivers. A three-layered feed forward artificial neural network was developed to predict the maximum stream temperature of Degirmendere River for the five days ahead. Satisfactory results were achieved as the average prediction error turned out to be less than 1°C

Stream temperature determines the rate of the decomposition of organic matter and the saturation concentration of dissolved oxygen. Combined with industrial waste, stream temperature becomes a crucial parameter. Therefore, estimation of maximum stream temperature is very important, especially during summertime when the high temperatures may become dangerous for the habitat of rivers. A three-layered feed forward artificial neural network was developed to predict the maximum stream temperature of Degirmendere River for the five days ahead. Satisfactory results were achieved as the average prediction error turned out to be less than 1 degrees C.