2026-08-17T09:31:45Z https://drops.dagstuhl.de/oai

oai:drops-oai.dagstuhl.de:23844 2025-11-12T13:20:28Z ddc:004 open_access

Accommodating Space-Time Scaling Issues in GAM-Based Varying Coefficient Models Comber, Alexis Harris, Paul Brunsdon, Chris Spatial Analysis Spatiotemproal Analysis The paper describes modifications to spatial and temporal varying coefficient (STVC) modelling, using Generalized Additive Models (GAMs). Previous work developed tools using Gaussian Process (GP) thin plate splines parameterised with location and time variables, and has presented a space-time toolkit in the stgam R package, providing wrapper functions to the mgcv R package. However, whilst thin plate smooths with GP bases are acceptable for working with spatial problems they are not for working with space and time combined. A more robust approach is to use a tensor product smooth with GP basis. However, these in turn require correlation function length scale or range parameters (ρ) to be defined. These are distances (in space or time) at which the correlation function falls below some value, and can be used to indicate the scale of spatial and temporal dependencies between response and predictor variables (similar to geographically weighted bandwidths). The paper describes the problem in detail, illustrates an approach for optimising ρ and methods for determining model specification. Schloss Dagstuhl – Leibniz-Zentrum für Informatik Alexis Comber and Paul Harris and Chris Brunsdon 2025 Is Part Of LIPIcs, Volume 346, 13th International Conference on Geographic Information Science (GIScience 2025) InProceedings Text doc-type:ResearchArticle publishedVersion application/pdf doi:10.4230/LIPIcs.GIScience.2025.15 urn:nbn:de:0030-drops-238440 https://drops.dagstuhl.de/entities/document/10.4230/LIPIcs.GIScience.2025.15 eng https://creativecommons.org/licenses/by/4.0/legalcode