This study assesses Arctic cyclone characteristics and associated precipitation and surface wind speeds using an ensemble of regional climate model (GEMCLIM) simulations at 0.5° resolution for the 1981–2099 period following the RCP8.5 scenario. Comparison of GEMCLIM simulation with observations for current climate (1981–2010) suggests that GEMCLIM realistically reproduces the spatial and seasonal variation of Arctic cyclone frequency and intensity, and associated precipitation, for winter and summer. Clear added-value is found for several regions, compared to the driving data. The pressure-wind speed relationships for each region are reasonably reproduced and more extreme winds associated with increasing cyclone intensity are realistically simulated. In addition, the spatial and temporal variations of observed extreme cyclones are well captured. In future climate (2070–2099), the winter cyclone intensity and frequency, and associated precipitation, are projected to increase and decrease over the Aleutian and Icelandic Low regions, respectively. For summer, the projected changes are relatively smaller than those for winter and vary with region. Interestingly, significant decreases in cyclone contribution to total precipitation are found for northern Canada and Eurasia regions, despite increases in cyclone-related precipitation amount. This suggests stronger influence of mesoscale systems on precipitation compared to synoptic-scale systems. Enhanced pressure-wind speed relationships are projected for Arctic Canada and the Chukchi and East Siberian Seas. The increase of extreme cyclones during autumn is primarily related to sea ice loss during summer, while for winter, large-scale circulation changes (i.e. Arctic dipole) are mostly responsible due to strong sea ice loss in the central Arctic during autumn. This study demonstrates the added-value of dynamic downscaling with respect to Arctic cyclone characteristics and associated surface variables and provides useful insights regarding their future projections for use in risk assessment studies.

这项研究使用RCP8.5情景之后的1981-2099年期间，以0.5°分辨率对区域气候模型（GEMCLIM）进行了模拟，评估了北极的气旋特征以及相关的降水和地表风速。GEMCLIM模拟与当前气候观测值（1981-2010年）的比较表明，GEMCLIM真实地再现了冬季和夏季北极气旋频率和强度以及相关降水的空间和季节变化。与行驶数据相比，可以找到多个区域的明确增值。合理地再现了每个区域的压力-风速关系，并且现实地模拟了与旋风强度增加相关的更多极端风。此外，可以很好地捕获观测到的极端气旋的时空变化。在未来的气候（2070年至2099年）中，预计冬季风的强度和频率以及相关的降水将分别在阿留申群岛和冰岛低洼地区增加和减少。对于夏季，预计的变化相对小于冬季，并且随区域而变化。有趣的是，尽管与旋风有关的降水量有所增加，但加拿大北部和欧亚地区的旋风对总降水的贡献却显着下降。这表明与天气尺度系统相比，中尺度系统对降水的影响更大。预计加拿大北极地区，楚科奇海域和东西伯利亚海域将增强压力风速关系。秋季极端飓风的增加主要与夏季的海冰损失有关，而冬季，大规模的环流变化（即北极偶极子）是造成秋季中部北极海冰大量流失的主要原因。这项研究证明了相对于北极气旋特征和相关的表面变量而言，动态降尺度的附加值，并提供了有关其未来预测的有用见解，可用于风险评估研究。