The potential impact of wildfire smoke on Arctic cirrus formation is discussed based on lidar and radar observations during the winter half year of the 1-year MOSAiC (Multidisciplinary drifting Observatory for the Study of Arctic Climate) expedition. Aerosol and ice cloud observations were performed aboard the icebreaker Polarstern at latitudes > 85° N. Aged Siberian wildfire smoke polluted the tropopause region over the central Arctic during the entire winter half year of 2019-2020. The smoke particle surface area concentration at the tropopause was of the order of 5-15 μm2cm-3 and indicated considerably enhanced levels of aerosol pollution for more than 6 months. Numerous cirrus systems with cloud-top temperatures between -60 and -75 °C developed in the polluted upper troposphere. We analyzed all MOSAiC winter cirrus layers with respect to their geometrical and optical properties and a subgroup of 20 cirrus events with respect to their ice water content (IWC) and ice crystal number concentration (ICNC). In individual ice fallstreaks that are connected to individual ice nucleation events, ICNCs typically ranged from 1 to 10 crystals L-1 but were frequently also as high as 20-50 L-1; however, observations > 100 L-1 were rare. Three observational facts corroborate our hypothesis that smoke significantly influenced Arctic cirrus formation: (1) the occurrence of a long-lasting, persistent smoke pollution layer in the upper troposphere so that favorable conditions for heterogeneous ice nucleation on smoke particles were always given and, at the same time, homogeneous freezing of background aerosol was probably widely suppressed; (2) the high smoke particle surface area concentrations, which were high enough to significantly trigger ice nucleation on smoke particles (as shown in Part 2, the companion paper to this article; ); and (3) the frequently found maximum cirrus ice saturation ratios of 1.3-1.5, which point to the dominance of heterogeneous ice nucleation processes, initiated by inefficient ice-nucleating particles (INPs), as expected when aged smoke particles (i.e., organic aerosol particles) serve as INPs. The studies are continued in the simulation portion of this work (Part 2; Ansmann et al., 2025).