R for spatio-temporal handling of moving polygons

Data cubes are structures to store and analyse spatio-temporal data in raster and vector format. Typical examples of spatio-temporal vector data are weather stations collecting data over time, or administrative polygons where historical data is aggregated per zone. A less explored use case for data cubes are moving polygons. Example of moving polygons would be spatial representations of glacier retreat, emergence of volcanic lava flows or the changes of a city boundary over time. In this contribution, I introduce the handling of polygons that evolve and move over time using vector data cubes. The implementation in R makes use of the packages {stars} and {cubble} as ways to represent data in array and tabular formats. The advantage of vector data cubes in both formats is the ability to apply common array operations, but also tidy data wrangling techniques to explore and analyse data. Temporal analyses can be performed using packages like {tsibble}, while spatial analyses can be performed using {sf} methods. Further, more complex spatio-temporal analyses like change detection can be performed using {stampr}. Visualization techniques using {ggplot2} and {tmap} are also explored.