Comment on gmd-2022-14

One of the major obstacles for designing solutions against the imminent climate crisis is the scarcity of high spatio-temporal resolution model projections for variables such as precipitation. This kind of information is crucial for impact studies in fields like hydrology, agronomy, ecology and risk management. The currently highest spatial resolution datasets on a daily scale for projected conditions fail to represent complex local variability. We used deep learning (DL) based statistical downscaling (SD) methods to obtain daily 1 km resolution gridded data for precipitation in the Eastern Ore Mountains in Saxony, Germany. We built upon the well established climate4R framework, while adding modifications to its base-code and introducing skip connections based DL architectures, such as U-Net and U-Net++. We also aimed to address the known general reproducibility issues by creating a containerized environment with multi-GPU and TensorFlow’s deterministic operations support. The perfect prognosis approach was applied using the ERA5 reanalysis and the ReKIS (Regional Climate Information System for Saxony, Saxony-Anhalt, and Thuringia) dataset. The results were validated with the VALUE framework. The introduced architectures show a clear performance improvement when compared to previous SD benchmarks. Characteristics of the DL models configurations that promote their suitability for this specific task were identified, tested and argued. Full model repeatability was achieved employing the same physical GPU.