A new global turbulence forecast product for aeronautics

Aeronautical turbulence is a major risk for this sector with dozen of millions of dollars for companies, and more than 7000 hours of sick leaves. Turbulence suffered by planes has multiple sources such as CAT (Clear Air Turbulence) which is particularly dangerous because non detectable. Météo-France has developed a new diagnostic of turbulence through a common project between DSM/AERO and DR/CNRM.
This new diagnostic is a combination of predictors following the method developed by NCAR for the GTG (Graphical Turbulence Guidance) product. Each predictor detects different sources of CAT, such as vertical or horizontal wind shear or thermal vertical instability. The choice of the best predictors is made by a learning process using EDR (Eddy Dissipation Rate) observed by commercial flights over the United States of America. In order to produce a final diagnostic in EDR (OACI recommendation), each predictor computed in numerical model is converted into this unit under the hypothesis that each predictor follows the same law of distribution than the observed EDR.
This new diagnostic has been built on the global numerical model ARPEGE and validated over the first semester of 2019 with objective scores using observed EDR and case studies using reports from pilots. It will be available every 10 Flight Levels with a time step of one hour.
As turbulence causes (jet, convection) are subjected to forecast errors, future works should evaluate the contribution of ensemble forecasts to this diagnostic.

The joint document (in French) details the new diagnostic and presents results on some case studies.