Predicting turbulence by satellite and validating with in situ data: A full-scale analysis with the GOES-R Tropopause Folding Turbulence Product

Tropopause folding has a well-established link to clear air turbulence in the upper troposphere. Here we show the latest version of an empirical algorithm that models tropopause folds in three dimensions, based on gradients observed in the GOES water vapor channel. These modeled tropopause folds are used to predict clear air turbulence, and are shown to have significant skill when validated with commercial airline pilot reports as well as automated turbulence reports.

The algorithm is an Option 2 derived product for the GOES-R Algorithm Working Group. We will present the results of a validation of this product with an independent in-situ eddy dissipation rate (EDR) dataset over 28 months, with a sensitivity analysis that considers associated satellite gradients, aircraft incident angle, and time of year. The target accuracy of the product is the prediction of moderate or greater (MOG) turbulence with a 50% false alarm rate. We will also analyze the problem of normalizing 1-minute EDR data to the temporal and spatial scale of tropopause folds, which has a broad applicability to other projects that validate with aircraft EDR data.