García Pérez, Alfonso2024-05-202024-05-202022-08-25Print: 978-3-031-15508-6, Online: 978-3-031-15509-3https://doi.org/10.1007/978-3-031-15509-3_23https://hdl.handle.net/20.500.14468/15446The spatio-temporal variogram is the key element in spatiotemporal prediction based on kriging, but the classical estimator of this parameter is very sensitive to outliers. In this contributed paper we propose a trimmed estimator of the spatio-temporal variogram as a robust estimator. We obtain an accurate approximation of its distribution with small samples sizes and a scale contaminated normal model.We conclude with an example with real data.eninfo:eu-repo/semantics/closedAccessbook part