Severe high altitude aircraft turbulence on thunderstorm peripheries
The horizontal divergence at a storm's top may be so large that it alters the ambient flow, leading to conditions favorable for severe turbulence some distance from the primary convection. Unknown in these cases is whether the altered flow increases vertical wind shear sufficient to generate the turbulence or if the altered flow creates gravity waves which further modify the local environment to generate it.
To shed light on this question, we examined two high altitude severe turbulence encounters with data computed on high resolution numerical forecast models available at the National Centers for Environmental Prediction. Since these models successfully predicted the actual convection, we applied the ULTURB algorithm to these two cases to examine the flow characteristics even though ULTURB was created to predict upper level turbulence in low Rossby number situations typical of synoptic scale flows.
In the first case, near Saint Louis, Missouri, on 8 May 2009, the model data showed that a large mesoscale convective complex over Missouri's southern half produced an outflow jet greater than 130 knots at FL400 in St. Louis's vicinity. This lowered the Richardson number to less than 0.75 but not low enough to expect turbulence to develop. The Lighthill-Ford radiation computed by the ULTURB algorithm was very large indicating the likelihood that gravity waves were emanating from the storm. Thus, the ULTURB forecast was for severe turbulence near St. Louis at FL400.
In a sidebar, for this case we examine two Lighthill-Ford radiation terms not included in ULTURB. From scaling arguments we expect these terms to be important in high Rossby number flows of thunderstorm environments, and indeed they are about an order of magnitude larger than the synoptic scale terms in this case but in the same general location as the terms already in ULTURB. However, because the synoptic scale Lighthill-Ford radiation was already an order of magnitude larger than in typical synoptic scale cases, the additional terms did not improve the successful ULTURB turbulence prediction. Many more cases need to be documented to determine if these additional terms need to be included in a ULTURB algorithm for thunderstorm peripheries.
The second case, near New Orleans, Louisiana, on 2 April 2009, was a different situation. The thunderstorm complex blocked an already strong jet level flow. This blocking slowed the ambient flow instead of accelerating it as in the first case. The Richardson number in the storm-altered flow was greater than one but, with sufficiently high Lighthill-Ford radiation near New Orleans ULTURB successfully predicted severe turbulence on the storm's periphery.
We conclude that for these two cases while the storm-modified flow did lower the flows' Richardson numbers, it was likely the generation of gravity waves caused the severe turbulence.