In the Arctic, days with clouds containing laminations (layers), as identified by visual inspection of lidar data (McCullough, 2019), are strongly correlated with ground-level precipitation (McCullough, 2021). This provides insight into the cloud conditions and processes involved in the production of snow. We now present a computationally-inexpensive computer algorithm that (1) automatically identifies laminated cloud features in time-altitude arrays of lidar data and (2) classifies the sky scene according to the presence of laminated clouds. A discretized residual method is used.</p> <p>Algorithm validation uses the 2016-2019 dataset from McCullough (2019): 532 nm backscatter, up to 5 km altitude, measured by the Canadian Network for the Detection of Atmospheric Composition Change (CANDAC) Rayleigh-Mie-Raman lidar (CRL; 1 min, 7.5 m resolution) at Eureka, Nunavut (79.6 N, 85.6 W). Sky scene classification at daily resolution is replicated with accuracy = 0.86. Correlations with snow are nearly identical: r = 0.62 (algorithm) and r = 0.63 (visual).</p> <p>The computer algorithm is then employed at hourly resolution (r = 0.50), and for regions below and above 1 km separately. Correlations with snow are largest when laminations occur simultaneously above and below 1 km (r = 0.44), followed by laminations exclusively above 1 km (r = 0.28); there is only a very small correlation for laminations exclusively below 1 km (r = 0.09). Therefore, cloud laminations above 1 km altitude are key lidar indicators of the production of precipitating snow detected at ground level.