Abstract
A Neural Network (NN) modelling approach has been shown to be successful in calculating pseudo steady state time and space dependent Dissolved Oxygen (DO) concentrations in three separate reservoirs with different characteristics using limited number of input variables. The Levenberg–Marquardt algorithm was adopted during training. Pre-processing before training and post processing after simulation steps were the treatments applied to raw data and predictions respectively. Generalisation was improved and over-fitting problems were eliminated: Early stopping method was applied for improving generalisation. The correlation coefficients between neural network estimates and field measurements were as high as 0.98 for two of the reservoirs with experiments that involve double layer neural network structure with 30 neurons within each hidden layer. A simple one layer neural network structure with 11 neurons has yielded comparable and satisfactorily high correlation coefficients for complete data set, and training, validation and test sets of the third reservoir.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Explore related subjects
Discover the latest articles, news and stories from top researchers in related subjects.References
Vollenweider RA (1975) Input-output models with special reference to the phosphorus loading concept in limnology. Schweitz Z Hydrol 37:53–84
Vollenweider RA (1976) Advances in defining critical loading levels for phosphorus in lake eutrophication. Mem 1st Ital Idrobiol 33:53–83
Carlson RE (1977) A trophic state index for lakes. Limnol Oceanogr 23:361–369
Rast W, Lee GF (1978) Summary analysis of the North American (US portion) OECD eutrophication project: nutrient loading-lake response relationship and trophic state indices. Environmental Research Series, No:EPA-600/3–78–008, US EPA, Environmental Research Laboratory, Oregon, USA
Ryding SO (1980) Monitoring of inland waters: OECD eutrophication programme- the Nordic project publication. Nordic Co-operative Organisation of Applied Research (NORDFORSK), Helsinki, Finland
OECD Study (1982) Eutrophication of waters, monitoring, assessment and control, final report. OECD, Paris
Vollenweider RA, Janus LL (1982) Statistical models for predicting hypolimnetic oxygen depletion rates. Mem Ist Ital Idrobiol 40:1–24
ASLO (1989) Predicting changes in hypolimnetic oxygen concentration with phosphorus retention, temperature, and morphometry. Limnol Oceanogr 34(7): 1359–1366
CE-QUAL-R1 (1986) A numerical one-dimensional model of reservoir water quality; user’s manual, instruction report E-82–1. Revised edition, U.S. Army Engineer Waterways Experiment Station, Vicksburg, Miss
CE-QUAL-W2 (1986) A numerical two-dimensional, laterally averaged model of hydrodynamics and water quality; user’s manual, instruction report E-86–5, U.S. Army Engineer Waterways Experiment Station, Vicksburg, Miss
Scardi M (1996) Artificial neural networks as empirical models for estimating phytoplankton production. Mar Ecol Ser 139:289–299
Recknagel F, French M, Harkonen P, Yabunaka K (1997) Artificial neural network approach for modeling and prediction of algal blooms. Ecol Modeling 96:11–28
Yabunaka K, Hosomi M, Mukarami A (1997) Novel application of a back-propagation artificial neural network model formulated to predict algal bloom. Wat Res 36:89–97
Karul C, Soyupak S, Germen E (1998) A new approach to mathematical water quality modeling in reservoirs: neural networks. Int Rev of Hydrobiology 83:689–696
Karul C, Soyupak S, Cilesiz AF, Akbay N, Germen E (2000) Case studies on the use of neural networks in eutrophication modeling. Ecol Modeling 134:145–152
Soyupak S, Yazıcı A, Cilesiz AF, Akbay, N (2001) A neural network based model for estimating dissolved oxygen in a very deep reservoir. 16th Int Symp on Comp and Inf Sci, ISCIS XVI, Antalya, Turkey
Soyupak S, Yemisen D, Mukhallalati L, Erdem S, Akbay N, Yerli S (1999) The spatial and temporal variability of limnological properties of a very large and deep reservoir. J Int Rev Hydrobiology 83:183–190
Soyupak S (1992) Keban baraj gölü ve havzası çevre sorunları projesi. TUBITAK–DEBAG 1/G Raporu, Ankara, Turkey, (in Turkish)
Soyupak S (1993) Keban baraj gölü ve havzası çevre sorunları projesi, TUBITAK –DEBAG 83/G Raporu, Ankara, Turkey, (in Turkish)
Yemisen D (1994) Keban baraj gölü ve havzası çevre sorunları projesi—final raporu—Keban baraj golu sularının fiziksel, kimyasal ve biyolojik özellikleri-Keban’da otrofikasyon- hidrodinamik modelleme sonuçları ve su kalitesi modelleme sonuçları- çozum onerileri, TUBITAK DEBAG 124/G Projesi, ODTU-DSI Genel Müdürlügü ve DSI 9. Bölge, Elazıg, Turkey, (in Turkish)
Hornik K, Stinchcombe M, White H (1989) Multilayer feedforward networks are universal approximators. Neural Networks 2(5): 359–366
Haykin S (1996) Adaptive filter theory, 3rd edn. Prentice Hall, Upper Saddle River, NJ
Hagan MT, Menjah M (1994) Training feed-forward networks with the Marquardt algorithm. IEEE Trans on Neural Networks 5(6):989–993
Ham FM, Kostanic I (2001) Principles of neurocomputing for science and engineering, McGraw-Hill International Edition, Electrical Engineering Series, NY
Eckman FH, Freeman WJ (1986) The sigmoid nonlinearity in neural computation: An experimental approach. In: Decker JS (ed) Neural computation, New York, American Institute of Physics, pp 135–145
Demuth H, Beale M (2000) Neural network toolbox user’s guide, The MathWorks Inc., Natick, MA
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Soyupak, S., Karaer, F., Gürbüz, H. et al. A neural network-based approach for calculating dissolved oxygen profiles in reservoirs. Neural Comput & Applic 12, 166–172 (2003). https://doi.org/10.1007/s00521-003-0378-8
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00521-003-0378-8