This special issue contains five papers. The first paper by E. Meerschman et al. (2013 — this issue) focuses on bivariate multiple-point statistics applied to proximal soil sensor data with the aim of mapping fossil ice-wedge polygons. Multiple-point statistics as a collection of geostatistical simulation algorithms which use a multiple-point training image as a structural model instead of a two-point variogram allows the simulation of more complex random fields. The paper shows the first application of multivariate multiple-point statistics in soil science. It outperformed the more traditional methods such as ordinary kriging and fuzzy k-means classification. It thus represents a new and promising tool for soil and soil sensor data processing.