Recent developments of functional kriging

The continuous improvement of technology have allowed the high frequency recording and storing of data measurements. Considering those data for each observation not any more as vector of points seen as fnite set of points but as functions seen as infinite set of points and applying statistical techniques on it, is the core concept of functional data analysis. So several functional versions of a wide range of classical statistical tools are being proposed and developed since the rise of functional data analysis as an important branch of statistics for the analysis of curves. Therefore in spatial statistics different functional versions of kriging predictor have been proposed according to the type of spatial data observed. Here we are mainly interested in the case of geostatistical data observed as functions. This paper provides an overview of how classical geostatiscal tools have been extended in order to deal with spatial prediction of georeferenced data observed or approximated as functions or curves.