Discovering variable phenomena by blowing up the data

Image-domain follow-up of transient and variable sources has been a huge success story for MeerKAT, resulting in many publications, including numerous high-impact ones. However, such sources could have been initially discovered by MeerKAT itself if a system for routinely creating a time-series of images was in place. The discovery and false positive rejection potential is expanded further if the full dimensionality of the data is explored, namely the spectral and polarimetric information that the telescope naturally provides in addition to the temporal and spatial dimensions. The computational cost of this can of course be quite significant. I will present some ideas related to this, framing the presentation around some archival MeerKAT observations of some globular clusters.