Conditional simulation and model-conditional uncertainty

Conditional Gaussian realizations require a retained kriging trend, variogram, conditioning data, and prediction grid. ps_surface_uncertainty() records the seed, simulation count, model, random-path behavior, and nugget treatment. The result is conditional on the covariance model and Gaussian assumptions; it does not include all measurement, conceptual-model, datum, or boundary uncertainty.

u <- ps_surface_uncertainty(ok_result, approach="conditional_simulation",
                            nsim=200, seed=17, keep_realizations=TRUE)