SOIL Studies

SOIL Studies

2017, Vol 6, Num, 3     (Pages: 001-007)

Determination of the Applicability of the Cokriging Method in Estimating Field Capacitiy

Tülay TUNÇAY 1 ,Oğuz BAŞKAN 1 ,İlhami BAYRAMİN 2 ,Şeref KILIÇ 3 ,Orhan DENGİZ 4

1 Toprak Gübre ve Su Kaynakları Merkez Araştırma Enstitüsü, Ankara
2 Ankara Üniversitesi Ziraat Fakültesi, Toprak ve Bitki Besleme Bölümü, Ankara
3 Ardahan Üniversitesi Mühendislik Fakültesi, Çevre Mühendisliği Bölümü, Ardahan
4 Ondokuz Mayıs Üniversitesi Ziraat Fakültesi, Toprak ve Bitki Besleme Bölümü, Ankara

DOI: 10.21657/topraksu.338299
Viewed: 1139
 - 
Downloaded : 512

In recent years, geostatistic has become a commonly used method in soil science. The co-kriging technique is widely utilized in determining the spatial distribution of particularly labor- and effortintensive analysis results. This study was conducted over a 32-km2 area (eight plots) within the Altınova Agricultural Enterprise, a subsidiary of the General Directorate of Agricultural Enterprises (TİGEM). A total of 135 soil samples (0–20 cm) were collected on a 500 m-spaced grid, and it was determined to soil texture in all soil samples. Additionally, field capacity (FC) values were determined through reduced sampling with an aim to assess the effectiveness of the co-kriging method in the estimation of FC values that require significant amounts of time and labor. On the basis of the results of correlation analysis, the clay value (R: 0.68) showing the highest correlation with FC was selected as a sub-variable. The effectiveness of the co-kriging technique in estimating the spatial structure of the FC value was tested with reduced proportions of FC values by taking the Mean Absolute Error (MAE) and Mean Squared Error (MSE) values into account. The results obtained highlighted the significant advantages of the co-kriging method in the estimation of soil properties. The co-kriging technique enabled accurate mapping of the spatial structure even for 63 FC values with low MAE and MSE values. The current study has shown that a high correlation coefficient is insufficient in accurate representation of the spatial structure by itself, but it requires an auxiliary variable.


Keywords : Cokriging, geostatistic, kriging, semivariogram