Our first outdoor tests of drifting around small-scale models (1975) showed great promise. Accurate results depend on two conditions. The surface upwind of the model must be very smooth, and wind speed must be just above the threshold speed, where drifting starts (Tabler, 1980b). With a frozen lake for our surface, model scales as small as 1-to-30 gave accurately scaled snowdrifts. Full-scale snowdrifts built by an entire winter of blizzards could be simulated in about two hours.
The models help predict complicated drift patterns. One of the first models simulated an array of cinderblocks in a test of snowdrifts in rough surfaces. Wind is from the upper right, over Ron's shoulder. The wisps of drifting snow are typical of conditions when scaling is accurate.