ABSTRACT: Inherent to a spatial variable is the scale at which it is measured. In many studies, variables are observed at different scales. For example, biomass data might be available at an aggregated level while temperature is usually measured at specific points. One of the main aims of statistical analysis is to make meaningful predictions, the accuracy of which may be achieved by including related variables in the model. The implementation of this may become cumbersome when related data are measured at different scales, particularly when the scales do not have a hierarchical structure. Currently, cokriging, the use of one or more spatial variables to predict another variable, is applied to variables of the same scale. In this work, we extend cokriging for use with variables of different scales by constructing a nonparametric cross-covariance matrix. This method is flexible as it applies to any marginal spatial model and is suited to large datasets because it uses latent variables which can assist with dimension reduction. The proposed nonparametric method is demonstrated with two correlated variables, biomass and temperature, which are measured at different spatial units. The results show that the method is appropriate for predicting data of different scales and that it outperforms some competing methods with respect to predictive performance.
Monday April 9, 2018 5:30pm - 7:00pm CDT
Monroe Room