Welcome | Instrument | Research | Data | Publications | Team | Links

Retrieval techniques

Radar-lidar : VarCloud (Delanoë et al 2008)

Varcloud is synergistic variational method which allows one to retrieve profiles of visible extinction coefficient, ice water content and effective radius in ice clouds using the combination of airborne or spaceborne radar, lidar and infrared radiometer. The forward model includes effects such as non-Rayleigh scattering by the radar and molecular and multiple scattering by the lidar. By rigorous treatment of errors and a careful choice of state variables and associated a priori estimates, a seamless retrieval is possible between regions of the cloud detected by both radar and lidar and regions detected by just one of these two instruments. Thus, when the lidar signal is unavailable (for reasons such as strong attenuation), the retrieval tends toward an empirical relationship using radar reflectivity factor and temperature, and when the radar signal is unavailable (such as in optically thin cirrus), accurate retrievals are still possible from the combination of lidar and radiometer.

Radar-only : RadOn (Delanoë et al 2007)

The RadOn stands for RADar ONly and can be used to characterise ice cloud properties when LNG lidar is not available. The method makes use of two measurements from a Doppler cloud radar, radar reflectivity and Doppler velocity, to retrieve ice crystal effective radius, ice water content, and visible extinction from which the optical depth can be estimated. This radar method relies on the concept of scaling the ice particle size distribution.

The Doppler velocity cannot be directly used by RadOn and we need to extract the vertical air motion (Protat and Williams 2011). This is a challenging task when RadOn is applied to airborne data, however this is overcome using the WIRE technique (Papazzoni 2010).

We have developed a new version of Radon, named RadonVar, which uses the variational approach similar to VarCloud.

Key point of both methods: The normalised particle size distribution approach (Delanoë et al. 2005).