Sensitivity of Satellite Lidar-Radar Cirrus Retrievals to PSD Assumptions: DARDAR-Nice v2 and Simulator

Ice clouds constitute a challenge to satellite remote-sensing due to the variability of their microphysical properties. A central parameter to understand and represent ice clouds in modelling as well as in remote-sensing is the ice particle size distribution (PSD), whose shape largely varies depending on the environmental conditions in which the ice cloud has formed and evolved. This shape is typically assumed in satellite retrieval algorithm, for instance as a mono-modal gamma-modified distribution. Our representation of PSDs has greatly improved over the last decades, largely due to novel parameterisation methods as well as the increasing availability and accuracy of in-situ measurements that can serve as a solid basis to calibrate retrieval algorithms.

This study investigates the impact of the PSD shape assumptions on cirrus retrievals obtained from lidar-radar satellite observations (DARDAR-Nice), with a strong focus on the ice crystal number concentration. Recent in-situ measurements from the JULIA dataset were recently processed to propose new parameterisations of the PSD that offer a better representation of small ice concentrations. The added-value of considering the observed bi-modality when representing PSDs for remote sensing applications was also discussed. We here assess the consequences of including such new parameterisations in DARDAR-Nice. Comparisons between v1 and v2 (offering updated PSD assumptions) of this satellite product are also discussed.

Finally, preliminary results from a DARDAR-Nice simulator will be shown. This simulator allows to perform synthetic lidar-radar observations and retrievals on high-resolution cloud model outputs. Comparisons between the model “truth” and synthetic retrievals will be investigated and discussed in the context of underlying PSD assumptions.