Title: Complex-valued covariance models for spatio-temporal vector data
Authors: Sandra De Iaco - University of Salento (Italy) [presenting]
Abstract: In geostatistical literature, there are various contributions focused only on modeling the spatial evolution of vector data with two components in the framework of the theory of complex-valued random fields. However, the temporal perspective is analyzed separately or used to model time-varying complex covariance models. Thus, in this context, it is surely challenging to propose some advances in modeling the joint spatial and temporal behavior of vector data with a reasonable representation on a complex domain. After introducing the theoretical background regarding the complex formalism of a spatio-temporal random field, some techniques for building new families of spatio-temporal models are discussed. Then, the spatio-temporal complex modeling is applied to sea current data referred to the US East and Gulf Coast. The results regarding a comparative analysis between different complex-valued covariance models are also presented.