Regional Assessment of Monthly Soil Temperatures in the Aegean Region of Turkey


Bilgili M.

ARABIAN JOURNAL FOR SCIENCE AND ENGINEERING, vol.37, no.3, pp.765-775, 2012 (SCI-Expanded) identifier identifier

  • Publication Type: Article / Article
  • Volume: 37 Issue: 3
  • Publication Date: 2012
  • Doi Number: 10.1007/s13369-012-0199-0
  • Journal Name: ARABIAN JOURNAL FOR SCIENCE AND ENGINEERING
  • Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus
  • Page Numbers: pp.765-775
  • Keywords: Artificial neural network, Air temperature, Regional assessment, Soil temperature, Solar energy, ARTIFICIAL NEURAL NETWORKS, SURFACE-TEMPERATURE, PREDICTION, SEDIMENT
  • Çukurova University Affiliated: Yes

Abstract

An artificial neural network (ANN) model was developed to predict monthly soil temperatures in the Aegean Region of Turkey. The model used soil temperature data measured by the Turkish State Meteorological Service between 2000 and 2006 at the Kutahya, Manisa, Usak, Afyonkarahisar, Izmir, Aydm, Denizli and Mugla meteorological stations at depths of 5, 10, 20, 50 and 100 cm below the surface. The monthly air temperature, depth and month of the year were used in the input layer of the artificial neural network model, while the monthly soil temperature was the target data in the output layer. A prediction model was developed using the MATLAB program and the results derived from the model were compared with measured values. The mean absolute error (MAE) for the derived results from the different depths at all stations ranged from 0.32-0.82 degrees C, while the corresponding range for the validation data set was 0.27-0.76 degrees C.