View article

The article approaches some problems related to the in influence of the space scale on the quality of statistically based spatial models of meteo-climatic variables. The neighborhood information issue is discussed, meaning the extent to which the information from the area surrounding a point of measurement is more relevant for spatialisation than the strictly local information. Then we approach the problem of spatialisation in heterogeneous regions and the applicability of the models at different scales. Finally, we tackle the problem of the outliers, the presence of which could lead us to incorrect interpretations.

The elaboration of statistically based spatial models for meteo-climatic variables requires insight on some aspects, which could help us improve the models or avoid errors. We discuss in this article some of these aspects, related to the influence of the space scale on the quality of the spatialisations.