Premium
NON‐PARAMETRIC ESTIMATION OF THE SILL IN GEOSTATISTICS
Author(s) -
CHERRY STEVE
Publication year - 1997
Publication title -
environmetrics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.68
H-Index - 58
eISSN - 1099-095X
pISSN - 1180-4009
DOI - 10.1002/(sici)1099-095x(199701)8:1<13::aid-env232>3.0.co;2-y
Subject(s) - variogram , geostatistics , mathematics , statistics , sill , parametric statistics , estimator , variance (accounting) , spatial dependence , econometrics , spatial variability , kriging , geology , accounting , geochemistry , business
A non‐parametric method of estimating the variance (sill in geostatistics) of spatial random variables of second‐order stationary processes is described. The estimator is the sum of the estimated parameters from a least squares fit of a linear combination of valid semivariograms to a sample semivariogram with the constraint that all parameters be non‐negative. Initially, the estimates tended to be biased and highly variable. A simple penalized fitting routine is described that yields less biased estimates with smaller variance. The method is compared with variance estimates determined from non‐linear fits of three parametric semivariogram models to sample semivariograms computed from simulated random fields. © 1997 by John Wiley & Sons, Ltd.