The impact of interpolation method parameters on the accuracy of predicted values
2013
Vyciene, G., Aleksandras Stulginskis Univ., Akademija, Kauno reg. (Lithuania) | Grybauskiene, V., Aleksandras Stulginskis Univ., Akademija, Kauno reg. (Lithuania)
Kriging is a method of interpolation, which predicts unknown values from data observed at known locations. This method uses a variogram to express spatial variation, and it minimizes the error of predicted values which are estimated by spatial distribution. The objective of the current work was to investigate how the accuracy of predicted runoff average depth values ranges by applying ordinary Kriging interpolation method when parameters of the method are altered. The simulation was tested by the following method parameters on: fitted semivariogram model, lag size, neighbouring points; the number of sectors of the circle. The best results of the spatial distribution of runoff average depth were received applying the ordinary Kriging method when the exponential variogram is used, the lag size being 27 km while 6 contiguous points and the orbicular scheme divided into sectors in 45° angle are taken. Applying the Ordinary Kriging method having selected all parameters with probability p = 0.95 the model described about 74% of all investigated values.
Показать больше [+] Меньше [-]