| Abstrakt: |
The goal of this research is to model the level of carbon dioxide flowing from soil to sky using various methods. The methods of multiple linear regression (MLR) and artificial neural networks (ANN) beside two different hybrid models were exploited to achieve this objective. These hybrid models were arranged as the prior two methods with principal component analysis (PCA). For the ANN, 36 different structures were used with different transfer (logsig–logsig, tansig–tansig, pureline–pureline, logsig–tansig, logsig–pureline and tansig–pureline)—learning functions (Levenberg–Marquardt and Gradient Descent with Momentum) and neuron numbers (10, 20 and 30). The manure norm, soil type, soil temperature, soil moisture content, soil depth, and photosynthetically active radiation values were taken into account as input parameters while CO2 flux was output parameter. According to the research conducted, the best results were obtained from the ANN method. This method was followed by PCA + ANN, MLR and PCA + MLR methods. The R2 value of the network established in the ANN method was determined as 0.98. In this ANN model, Levenberg–Marquardt and tansig–pureline with 30 neurons were used as transfer and learning functions, respectively. Besides, when principal components were used as input parameters, the lower R2 values were obtained with both the MLR and ANN methods. [ABSTRACT FROM AUTHOR] |
Full text from SpringerLink