Within the ongoing DWD-KIT cooperation on radar data correction and merging, we present a new framework for evaluating precipitation merging methods. We implemented a parallelized leave-many-out validation setup to systematically test several adjustment methods available in mergplg, an open-source collection of such methods. The focus is on benchmarking existing merging techniques before developing novel approaches, including AI-based methods, to improve quality and temporal resolution in adjusted radar precipitation fields.

Please feel free to share this invitation with interested colleagues at KIT.