Due to data availability long-term variations in precipitation rates are mostly studied based on daily precipitation recordings. Recent research suggests, however, that variations in sub-daily precipitation are subject to higher dynamics compared to daily precipitation and a more rapid intensification is likely. Here we show that both observational data with at least 58 years of sub-daily precipitation records and a dynamical downscaling approach with low spatial resolution based on atmospheric re-analysis data confirm these expectations with consistent results. High percentiles of precipitation are subject to multi-decadal oscillations and increased during the last 150 years. We found an increase of 4% K⁻¹ (daily), 12% K⁻¹ (hourly), and 13% K⁻¹ (10 min), which is consistent with Clausius–Clapeyron- (CC) and super CC-scaling, respectively. These findings highlight that dynamical downscaling can help to reliably shed light on sub-daily precipitation variations if small timescales are considered in the experiments.

由于数据的可用性，大部分降水率的长期变化是根据每日的降水记录来研究的。但是，最近的研究表明，与日降水量相比，次日降水量的变化具有更高的动态性，并且集约化速度可能更快。在这里，我们表明，具有至少58年的次日降水记录的观测数据，以及基于大气再分析数据的低空间分辨率的动态降尺度方法，都证实了这些期望并具有一致的结果。高降水百分位会经历十年的振荡，并在过去150年中增加。我们发现每天增加4％K ^-1，每小时增加12％K ^-1和13％K ^-1（10分钟），这分别与Clausius–Clapeyron-（CC）和超级CC缩放一致。这些发现表明，如果在实验中考虑较小的时间尺度，则动态降尺度可以帮助可靠地揭示次日降水量的变化。