sptotal: an R package for predicting totals and weighted sums from spatial data

Area of Science:

• Ecology
• Environmental Science
• Spatial Statistics

Background:

• Ecological and environmental surveys often require predicting population means or totals within finite regions.
• Complete region sampling is frequently infeasible due to time and financial constraints.

Purpose of the Study:

• To introduce the sptotal R package, offering software for predictions in ecological surveys.
• To provide accurate predictions and associated standard errors for quantities of interest in finite populations.

Main Methods:

• Utilizes the Finite-Population-Block-Kriging (FPBK) predictor, a method that accounts for spatial correlation.
• Incorporates specific variance reduction techniques suitable for finite population sampling.

Main Results:

• The sptotal package enables prediction of population totals and means with associated uncertainty estimates.
• The FPBK predictor effectively handles spatial dependencies and finite population characteristics.

Conclusions:

• The sptotal R package offers a valuable tool for ecological and environmental researchers facing sampling limitations.
• The FPBK method provides a robust approach for spatial prediction in finite populations, enhancing survey accuracy.