This is a companion dataset to the paper "Precipitation Variability in CMIP6 Climate Models across the North Atlantic–European region and their Links to Atmospheric Circulation" by E. Plavcová submitted to Climate Dynamics.

Long-term changes in climate variability represent an important aspect of climate change, with various impacts on society and environment. In this study, we analyze outputs from 13 CMIP6 global climate models (GCMs) across the North Atlantic–European domain, focusing on their simulations of precipitation probability and short-term variability in both historical and future climates. Precipitation probability denotes the probability of a wet day (> 1 mm), and precipitation variability reflects the tendency to cluster wet days into sequences. By comparing against the ERA5 reanalysis, we found that the GCMs tend to overestimate precipitation probability across Europe in winter, whereas in summer, they have a tendency to underestimate it around 50°N. Precipitation variability is, on average, underestimated by the GCMs in summer, while overestimated in several regions in winter. Projections for the end of the 21st century indicate significant changes in both precipitation probability and variability which are more pronounced under the more pessimistic emission scenario compared to the moderate one. We found that the changes in probability and variability are mutually independent: the former being more latitudinal-dependent while the latter differs between the west and east. After identifying atmospheric circulation conducive and non-conducive to precipitation occurrence, we found that GCMs overestimating the frequency of conducive circulation tend to overestimate precipitation probability, and vice versa. Furthermore, increased precipitation variability is associated with higher circulation variability. Finally, our analysis reveals that projected changes in precipitation probability and variability are often linked to projected changes in atmospheric circulation, especially in winter.气候变率的长期变化是气候变化的重要组成部分，对社会与自然环境均产生广泛影响。本研究针对北大西洋-欧洲区域，分析了13个第六次国际耦合模式比较计划（CMIP6）全球气候模型（Global Climate Models，GCMs）的模拟输出结果，重点关注其在历史气候与未来气候情景下对降水概率及短时变率的模拟表现。其中，降水概率指单日降水量≥1毫米的湿日发生概率，降水变率则表征湿日聚集为连续时段的趋势。通过与欧洲中期天气预报中心第五代再分析资料（ERA5）对比，本研究发现：冬季时，上述GCMs对欧洲全域的降水概率均存在高估现象；而夏季时，其在北纬50°附近区域则呈现低估特征。夏季，GCMs对降水变率的模拟整体偏低；冬季则在部分区域存在高估情况。针对21世纪末的气候预估结果显示，降水概率与降水变率均将发生显著变化，且相较于中等排放情景，在更悲观的排放情景下，此类变化更为突出。研究表明，降水概率与降水变率的变化相互独立：前者更具纬向依赖性，而后者则呈现东西区域差异。在甄别出有利于与不利于降水发生的大气环流类型后，研究发现：GCMs若高估了有利环流的发生频次，则其对降水概率的模拟也会偏高，反之亦然。此外，降水变率的增强与环流变率的升高呈正相关关系。最终，本研究分析结果表明，预估的降水概率与降水变率变化，通常与大气环流的预估变化存在关联，这一特征在冬季尤为显著。