Lesson 29f Kriging
Boston University EE509 "Applied Environmental Statistics" Course: The sixth lecture in our unit on spatial statistics covers Kriging in greater detail, presenting the steps for Maximum Likelihood fitting a spatial covariance functions in R, selection among alternative covariance functions (AIC), polynomial detrending, and using the correlation function to interpolate both a mean and standard error. We also briefly discuss the limitations and assumptions of Kriging (e.g. ignored parameter uncertainties, stationarity, isotropy, information lost from fitting as a multi-step process) and foreshadow the application of spatially-correlated data models to any process model. http://people.bu.edu/dietze/Bayes2020...