|
|
|
|
|
| International Agrophysics |
| wydawca: | Instytut Agrofizyki
im. B. Dobrzańskiego
PAN
w Lublinie |
| ISSN: |
0236-8722 |
vol. 25, nr. 2 (2011)
|
|
|
poprzedni artykuł wróć do listy artykułów następny artykuł
|
|
|
Forecasting nitrate concentration in groundwater using artificial neural network and linear regression models
|
|
| (pobierz wersję PDF )
|
|
|
A.H. Zare1, V.M. Bayat1, A.P. Daneshkare2
|
|
|
1 Agricultural Irrigation and Drainage Engineering Department, Faculty of Agriculture, Bu-Ali Sina University, Hamedan, Iran |
|
|
2 Agricultural Irrigation and Drainage Engineering Department, Faculty of Agriculture, Ghazvin University, Ghazvin, Iran |
|
|
vol. 25 (2011), nr. 2,
pp. 187-192
|
|
|
streszczenie
In this research, the ability of artificial neural network (ANN) to model groundwater nitrate of Arak Aquifer (Iran) is introduced. The ANN and linear regression (LR) methods were used to relate groundwater nitrate concentration to other water quality indices. Results showed that using the measured parameters is convenient to model nitrate concentration with acceptable and appropriate accuracy and ANN and LR methods were able to predict nitrate concentration at the desirable level of accuracy. Comparison of ANN analysis with LR model results showed that ANN requires fewer parameters with more accuracy in comparison to LR models. However, the ANN model with the highest correlation coefficient (r = 0.87), minimum root mean square error (RMSE=10.46 mg l-1) and mean absolute error (MAE = 7.77 mg l-1) provided the best results among the LR models.
|
|
słowa kluczowe
nitrate concentration, prediction, artificial neural network, regression
|
|
|
|
|
|
|
