SOIL Studies

SOIL Studies

2019, Vol 8, Num, 3     (Pages: 115-121)

Modeling of Soil Temperature in Central Black Sea Region by Time Series Analysis Method

Filiz KARA 1 ,Bilal CEMEK 2

1 Batı Akdeniz Tarımsal Araştırma Enstitüsü, Antalya
2 Ondokuz Mayıs Üniversitesi, Ziraat Fakültesi, Tarımsal Yapılar ve Sulama Anabilim Dalı, Samsun

DOI: 10.21657/topraksu.655501
Viewed: 980
 - 
Downloaded : 390

The aim of this study is to estimate the monthly average soil temperature measured in different layers (5, 10, 20, 50, 100 cm) of the Central Black Sea Region by Box Jenkins (BJ) technique. The monthly average soil temperature values of the years 1969-2013 were analyzed and seasonally adjusted and then the models suitable for these datasets were determined. The projections for 2013-2020 were obtained. The appropriate model for each layer of Samsun, Amasya and Ordu is determined as seasonal autoregression. The determination coefficient (R2) values of these stations are between 0.59-0.96, the standard deviation (RMSE) of the estimation error is between 1.55-4.93 and the mean absolute error (MAE) values are 1.24-4.28 values have changed. The appropriate model for each layer of Tokat and Çorum was determined as Seasonal Moving Average. The coefficient of determination (R2) of these stations is between 0.83-0.95 and the standard deviation (RMSE) of the estimation error is between 1.79-4.68 and the mean absolute error (MAE) is 0.03-3.91 values have changed. As a result of the study, it was seen that time series models yielded successful results in the monthly average soil temperature calculations in the Central Black Sea Region provinces.


Keywords : ARIMA models, box jenkins technique, climate, soil temperature