CAT algorithms using outputs of large scale numerical weather prediction models have been available for many years.

In the past years, with the increasing time and space resolutions of the numerical weather prediction models, the need of evaluating the current CAT algorithms within this evolving framework arises .

Within the Eumetnet E-AMDAR program, commercial aircrafts, using integrated sensors, measure meteorological parameters and downlink them in real time to the ground using AMDAR text messages. The AMDAR messages contain two parameters relevant to turbulence: a turbulence intensity and a vertical gust index.

At the end of 2005, a forty days AMDAR Intensive Observation Period (IOP) took place during which up to 500 000 turbulence observations were gathered over western Europe.

Before carrying out the intercomparison, some statistical analyses and quality controls are performed on the AMDAR data observations collected during the 2005 IOP . Then several CAT algorithms are calculated using the meso-scale numerical weather prediction model ALADIN from Météo France. The CAT algorithms output are then compared to the AMDAR IOP turbulence observations.