Tropopause folding at satellite‐observed spatial gradients: 2. Development of an empirical model

[1] We develop an empirical model that estimates the horizontal distribution of tropopause folds based primarily on a robust relationship between upper tropospheric specific humidity gradients and tropopause folding. The specific humidity gradients are resolved by the altered water vapor (AWV) product, developed from the GOES imager water vapor channel. Unexpectedly, no proportional relationship was found between tropopause fold size and AWV gradient magnitude beyond the one-to-one occurrence of strong gradients at tropopause folds. This indicates that tropopause folds do not commonly reach a “steady state” mass of ozone inside the fold. Rather, the mass of air must be controlled at least as much by the age of the fold and dynamical dispersion mechanisms that are not linearly related to the strength of the corresponding tropopause depression. This has important implications for the interpretation of lidar imagery and in situ ozone data at tropopause folds. In lieu of this result, a uniform fold size is assigned at all locations of modeled folding on the basis of the average of the folds used in parameterization. The model estimates the distribution of actively developing and residual tropopause folds, which is a useful method of tracking the dispersal of all stratospherically enhanced air in the troposphere and is unique to this observation-based model of tropopause folding.