Talk at GISRUK Conference
Our team member René Westerholt recently presented an approach for characterising heterogeneous spatial clusters at the 29th GIS Research UK Conference (Cardiff; this year held online). The approach is based on spatial eigenvector filtering (and a previous, recent paper from our group), which allows to selectively filter out unwanted spatial effects from data. In combination with a measure of spatial variance unevenness and a hotspot estimator, the application of this filtering allows a better understanding of internally fluctuating spatial clusters than with existing methods. The methodology is tested in the short paper using a novel indicator for "e-food deserts" from the University of Leeds with a focus on the UK West Midlands. It can be shown that an established older approach mainly highlights the often very heterogeneous boundaries of clusters, while the presented approach allows for a better characterisation of heterogeneous cluster centres.
The short paper can be found as follows:
Westerholt, R. (2021): Exploring and characterising irregular spatial clusters using eigenvector filtering. 29th Annual GIS Research UK Conference (GISRUK), Cardiff, UK. DOI: 10.5281/zenodo.4665575.
There is also a video available on YouTube: