University of Hertfordshire

Surface roughness during depositional growth and sublimation of ice crystals

Research output: Contribution to journalArticle


  • Cedric Chou
  • Jens Voigtländer
  • Zbigniew Ulanowski
  • Paul Herenz
  • Henner Bieligk
  • Tina Clauss
  • Dennis Niedermeier
  • Susan Hartmann
  • Georg Ritter
  • F. Stratmann
View graph of relations
Original languageEnglish
Number of pages16
Pages (from-to)13687-13702
JournalAtmospheric Chemistry and Physics
Journal publication date27 Sep 2018
Early online date18 Apr 2018
Publication statusPublished - 27 Sep 2018


Ice surface properties can modify the scattering properties of atmospheric ice crystals and therefore affect the radiative properties of mixed-phase and cirrus clouds. The Ice Roughness Investigation System (IRIS) is a new laboratory setup designed to investigate the conditions under which roughness develops on single ice crystals, based on their size, morphology and growth conditions (relative humidity and temperature). Ice roughness is quantified through the analysis of speckle in 2-D light-scattering patterns. Characterization of the setup shows that a supersaturation of 20 % with respect to ice and a temperature at the sample position as low as-40 °C could be achieved within IRIS. Investigations of the influence of humidity show that higher supersaturations with respect to ice lead to enhanced roughness and irregularities of ice crystal surfaces. Moreover, relative humidity oscillations lead to gradual ratcheting-up of roughness and irregularities, as the crystals undergo repeated growth-sublimation cycles. This memory effect also appears to result in reduced growth rates in later cycles. Thus, growth history, as well as supersaturation and temperature, influences ice crystal growth and properties, and future atmospheric models may benefit from its inclusion in the cloud evolution process and allow more accurate representation of not just roughness but crystal size too, and possibly also electrification properties.


Full version of an earlier discussion paper (Chou et al. 2018)

Research outputs


ID: 15270964