Some statistical problems related to singular spatial processes

Abstract

In the literature upon the matter the problem of singularity is practically neglected. The explanation that generally is given to this omission is that singular processes are too "pathological". I think this is a rather partial point of view. So in the paper I have tried to illustrate this position briefly examinating (sometime sending back to an appropriate bibliography) the ways choosen to solve the problem (even if the solution is still hard to find, as it is noted in the text). So I examine a way that should let some procedure of estimation. In the first and second paragraphs are used existing definitions and theorems, in the third paragraph some useful properties of the L-fields are developed. In fourth and fifth paragraph are exposed some procedures of estimations of tests. In the last paragraph the conclusions and the open problems are enclosed.