Ubiquitous Low-Level Surface Roughness of Cirrus-Like Ice

The shape and surface morphology of ice crystals comprising cirrus clouds are important for governing how these clouds re-direct sunlight, but are hard to study in situ. Use of scanning electron microscopy to investigate similarly-shaped crystals permits characterization of the surface morphology of these crystals with exceptionally high spatial resolution and statistical confidence. Here, we present evidence that points to an unexpected, ubiquitous low-level roughening on otherwise smooth-looking ice facets. Statistical analysis shows that this roughness is well-described by a Weibullian statistical model close to its Gaussian limit. The self-similar scaling inherent to this model motivates construction of a punctuated localized growth model for faceted ice crystal growth and ablation that yields surface morphologies that are also well-described by Weibullian statistics. We discuss merits and shortcomings of these findings, and their implications for atmospheric remote sensing.