This paper evaluates the effectiveness of Artificial Neural Networks (ANNs) for the estimation of the nitrate concentration in a study area located in the Nitrate Vulnerable Zone (NVZ) of the Arborea plain (Sardinia - Italy). Agricultural derived nitrate contamination of groundwater has been estimated by using easily and economical quantifiable parameters such as pH, electrical conductivity, temperature, groundwater level. Data used for training and validating the ANNs derive from a set of 225 measurements coming from 12 piezometers distributed in the study area. In order to define the best topology of the ANN and the best dimension of respectively the training and the validation sets a growing procedure has been applied.
Prediction of nitrate concentration in groundwater using an Artificial Neural Network (ANN) approach / Foddis, Ml; Montisci, A; Uras, G; Matzeu, A; Seddaiu, G; Carletti, A. - (2012). (Intervento presentato al convegno International Conference on Agricultural Engineering CIGR-AgEng2012 tenutosi a Valencia nel 8-12 July 2012).
Prediction of nitrate concentration in groundwater using an Artificial Neural Network (ANN) approach
Seddaiu G;Carletti A
2012-01-01
Abstract
This paper evaluates the effectiveness of Artificial Neural Networks (ANNs) for the estimation of the nitrate concentration in a study area located in the Nitrate Vulnerable Zone (NVZ) of the Arborea plain (Sardinia - Italy). Agricultural derived nitrate contamination of groundwater has been estimated by using easily and economical quantifiable parameters such as pH, electrical conductivity, temperature, groundwater level. Data used for training and validating the ANNs derive from a set of 225 measurements coming from 12 piezometers distributed in the study area. In order to define the best topology of the ANN and the best dimension of respectively the training and the validation sets a growing procedure has been applied.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
