CMStatistics 2021: Start Registration
View Submission - CMStatistics
B1349
Title: Directional spatial autoregressive dependence in the conditional first- and second-order moments Authors:  Miryam Sarah Merk - Georg-August-Universitaet Goettingen, Lehrstuhl fuer Oekonometrie/ Statistik (Germany) [presenting]
Philipp Otto - University of Glasgow (United Kingdom)
Abstract: In contrast to classical econometric approaches which are based on prespecified isotropic weighting schemes, we suggest that the spatial weighting matrix in the presence of directional dependencies should be estimated. In contrast to temporal autoregressive processes, where the direction of dependence is known in advance, we identify the direction of the spatial dependence based on different candidate neighbourhood sets. Two different types of processes, namely spatial autoregressive and spatial autoregressive conditional heteroscedastic processes, are considered and both estimated by maximum likelihood. Monte Carlo simulation results indicate that the models performance improves with sample size and with smaller neighbourhood subset sizes. Moreover, we apply the proposed approach to aerosol observations over the North Atlantic region and show that their spatial dependence matches the direction of the trade winds in this area.