Filtered Kriging for Spatial Data with Heterogeneous Measurement Error Variances

When predicting values for the measurement-error-free component of an observed spatial process, it is generally assumed that the process has a common measurement error variance. However, it is often the case that each measurement in a spatial data set has a known, site-specific measurement error variance, rendering the observed process nonstationary. We present a simple approach for estimating the semivariogram of the unobservable measurement-error-free process using a bias adjustment of the classical semivariogram formula. We then develop a new kriging predictor that filters the measurement errors. For scenarios where each site's measurement error variance is a function of the process of interest, we recommend an approach that also uses a variance-stabilizing transformation. The properties of the heterogeneous variance measurement-error-filtered kriging (HFK) predictor and variance-stabilized HFK predictor, and the improvement of these approaches over standard measurement-error-filtered kriging are demonstrated using simulation. The approach is illustrated with climate model output from the Hudson Strait area in northern Canada. In the illustration, locations with high or low measurement error variances are appropriately down- or upweighted in the prediction of the underlying process, yielding a realistically smooth picture of the phenomenon of interest.

[1]  D. Cox,et al.  An Analysis of Transformations , 1964 .

[2]  William F. Christensen,et al.  Dirichlet based Bayesian multivariate receptor modeling , 2008 .

[3]  Cecilia M. Bitz,et al.  Dynamics of Recent Climate Change in the Arctic , 2002, Science.

[4]  Lynne B. Hare,et al.  Statistical Principles of Research Design and Analysis , 1995 .

[5]  Carol A. Gotway,et al.  Statistical Methods for Spatial Data Analysis , 2004 .

[6]  P. Diggle.  Applied Spatial Statistics for Public Health Data , 2005 .

[7]  N. Cressie.  Fitting variogram models by weighted least squares , 1985 .

[8]  Richard G. Jones,et al.  A Regional Climate Change Assessment Program for North America , 2009 .

[9]  G. Matheron.  Principles of geostatistics , 1963 .

[10]  N. Cressie.  The origins of kriging , 1990 .

[11]  William F. Christensen,et al.  Pollution source apportionment using a priori information and positive matrix factorization , 2007 .

[12]  R. R. Hocking.  Methods and Applications of Linear Models: Regression and the Analysis of Variance , 2003 .

[13]  L. Waller,et al.  Applied Spatial Statistics for Public Health Data: Waller/Applied Spatial Statistics , 2004 .

[14]  Jack P. C. Kleijnen,et al.  Robustness of Kriging when interpolating in random simulation with heterogeneous variances: Some experiments , 2005, Eur. J. Oper. Res..

[15]  Noel A Cressie,et al.  Statistics for Spatial Data, Revised Edition. , 1994 .

[16]  William F. Christensen,et al.  Latent Variable Modeling for Integrating Output from Multiple Climate Models , 2012, Mathematical Geosciences.

[17]  J. S. Hunter,et al.  Statistics for Experimenters: An Introduction to Design, Data Analysis, and Model Building. , 1979 .

[18]  N. Cressie,et al.  Mean squared prediction error in the spatial linear model with estimated covariance parameters , 1992 .

[19]  C. Houser,et al.  Variations in Sea Ice in the Hudson Strait: 1971-1999 , 2003 .

[20]  Hirosi Hudimoto.  On the empirical bayes approach to classification in the case of discrete multivariate distribution having only finite mass points , 1979 .