A study of variance estimation methods for systematic spatial sampling

摘要: Abstract An undesirable property of systematic spatial sampling is that there no known method allowing unbiased estimation the uncertainty statistical estimates from these surveys. A number alternative variance methods have been tested and reported by various authors. Studies comparing estimators are inconclusive, partly because studies compare different sets estimators. In this paper, three recommended in recent compared using a single test dataset with properties. The first estimator study (ST4) based on post-stratification data. second (V08) predetermined correction factor calculated autocorrelation. third (MB) model prediction values semivariogram. MB ST4 were both found to be fairly accurate, while V08 consistently underestimated study. relies assumption autocorrelation structure can described particular exponential function. most likely explanation weak result for violated empirical data used experiment. better calculated, but safe approach use or ST4.