Forest inventory: further results for optimal sampling schemes based on the anticipated variance
Daniel Mandallaz, Adrian Lanz · Canadian Journal of Forest Research · 2001
This work presents optimal allocation rules for two-phase, two-stage sampling schemes in which the sampling density and the costs of the second phase can vary over domains. The optimality criterion is based on the anticipated variance. It also gives an improved version of discrete approximation for the resulting inclusion probabilities. An example illustrates the theory.