On the reflection and polarisation properties of ice cloud

A number of cirrus ice crystal scattering models are tested using measurements of total reflectance and polarised reflectance obtained from the space-based polarisation and directionality of Earth's reflectances (POLDER) instrument. In this paper, 1 day of global POLDER data is utilised taken from the 25 June 2003 to test the assumed ice crystal models. The POLDER instrument is able to test the validity of various ice crystal models since it can measure the total reflectance and polarised reflectance at up to 14 different viewing directions almost simultaneously between the scattering angles of about 60-180°. It is found that ice crystal models that are randomised (in this case the randomisation element is through distortion) from some pristine ice crystal geometry best fit simultaneous measurements of total and polarised reflectance. The optimal distortion parameter that best describes the POLDER measurements is found to be 0.40, which has been applied to a randomly oriented six-branched bullet-rosette and randomly oriented chain-like aggregate. Moreover, distorted ice crystals that have undergone significant distortions beyond 0.40 may fit the total reflectance measurements but not the polarisation measurements. Therefore, total reflectance measurements by themselves do not provide sufficient information to constrain assumed complex/distorted ice crystal models.