Neural network models as a management tool in lakes


Karul C., Soyupak S., Yurteri C.

HYDROBIOLOGIA, cilt.408, ss.139-144, 1999 (SCI-Expanded) identifier

  • Yayın Türü: Makale / Tam Makale
  • Cilt numarası: 408
  • Basım Tarihi: 1999
  • Doi Numarası: 10.1023/a:1017007313690
  • Dergi Adı: HYDROBIOLOGIA
  • Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED)
  • Sayfa Sayıları: ss.139-144
  • Anahtar Kelimeler: eutrophication, ecological modelling, multiple regression
  • Orta Doğu Teknik Üniversitesi Adresli: Hayır

Özet

A research was made on the potential use of neural network based models in eutrophication modelling. As a result, an algorithm was developed to handle the practical aspects of designing, implementing and assessing the results of a neural network based model as a lake management tool. To illustrate the advantages and limitations of the neural network model, a case study was carried out to estimate the chlorophyll-a concentration in Keban Dam Reservoir as a function of sampled water quality parameters (PO4 phosphorus, NO3 nitrogen, alkalinity, suspended solids concentration, pH, water temperature, electrical conductivity, dissolved oxygen concentration and Secchi depth) by a neural network based model. Alternatively, the same system was solved with a linear multiple regression model in order to compare the performances of the proposed neural network based model and the traditional linear multiple regression model. For both of the models, the linear correlation coefficients between the logarithms of observed and calculated chlorophyll-a concentrations were calculated. The correlation coefficient R, the best linear fit between the observed and calculated values, was evaluated to assess the performances of the two models. R values of 0.74 and 0.71 were obtained for the neural network based model and the linear multiple regression model, respectively. The study showed that the neural network based model can be used to estimate chlorophyll-a with a performance similar to that of the traditional linear multiple regression method. However, for cases where the input and the output variables are not linearly correlated, neural network based models are expected to show a better performance.