DETERMINATION OF THE SOIL TEMPERATURES IN AEGEAN REGION BY ARTIFICIAL NEURAL NETWORK METHOD

WOS: 000277481000013In this study, an artificial neural network (ANN) model developed in order to estimate the average monthly soil temperatures in the Aegean Region of Turkey based on the topographic data. The soil temperatures, which have been taken from Izmir, Mugla, Aydin, Denizli, Usak, Manisa, Kutahya and Afyonkarahisar meteorological stations, have been measured between the years 2000 and 2006 by the Turkish State Meteorological Service (TSMS), and temperatures have been measured at depths of 5, 10, 20, 50 and 100 cm respectively below the ground level. The latitude, longitude, altitude, depth and lunar information have been used in the input layer of the built artificial neural network model while the soil temperatures recognized as a target data have been used in the output layer. An estimation model has been developed employing MATLAB program and the derived results have been compared with the actual values. As a result, it has been seen that the error values have been found within the acceptable limits