Extended Abstract (1.4M)11.1
The Meteorological Environment Surrounding the Air France #447 Disaster
Michael Kaplan, DRI, Reno, NV; and D. R. Vollmer
Air France Flight #447 descended into the North Atlantic Ocean early on 1 June 2009, killing all on board. Contact with the Airbus A330 was lost when the flight was several hundred kilometers northeast of the South American coast at an altitude of ~12 km while the aircraft was in proximity to widespread deep convection, which was estimated via satellite to exceed 17 km. There has been much speculation but very little proof that weather played an important role in the incident as a result of icing, lightning, turbulence or some combination of these hazards. Based upon the final ACARS transmission as well as South American radiosonde, satellite, and GFS analysis/simulation data it is possible to reconstruct the horizontal and vertical atmospheric structure at the meso-α/β scale in proximity to the incident. These data lend one to speculate that a favorable environment for aviation turbulence may have existed at this time surrounding the deep convection, and we will describe some of these features in this preprint. Implicit in our analysis is related research by Vollmer (2008) and Kaplan et al. (2005) that serves as a theoretical/synoptic framework for the analyses presented here.
The larger scale environment during the 0000 – 0300 UTC 1 June 2009 period contained the following key synoptic features near the location of the incident (approximately 30ºW, 4ºN) as determined from the 0000 UTC Fernando de Noronha, Brazil sounding, GFS simulations, GFS analyses, and satellite-derived cloud track wind fields as well as infra-red and water vapor imagery: 1) an anticyclonically curved 100-150 mb subtropical jet streak oriented northwest-southeast in which the left exit region was approaching the incident location, 2) an anticyclonically-curved 70-100 mb tropical jet streak oriented southwest-northeast in which the left exit region was approaching the incident location, 3) a well-defined region of velocity convergence just to the northwest of the moist convection, which was occurring in proximity to and under velocity divergence over the convection and bracketing the incident location in the vertical, 4) substantial cross-jet frontogenesis due to 70 mb warming over and just downstream from the incident location, 5) increasing vertical wind directional shear in the layers including 150 – 100 mb and 100 – 70 mb primarily from northwest to north to southwest and 6) a dry adiabatic layer between 150 and 100 mb undercutting a strengthening inversion in time in between 100 and 70 mb. The inversion above 100 mb was also in proximity to the top of the moist convection within the anvil region.
These predominantly meso-α scale features are similar in many ways to the environment surrounding two mountain wave turbulence events analyzed by Vollmer (2008) and a much larger number of convective turbulence events analyzed by Kaplan et al. (2005). They are consistent with the superposition of two distinct tropopause boundaries, with an unstable layer above the lower tropopause capped by a stable layer above and within the upper tropopause. The wind direction and stratification accompanying the two jet streaks varied substantially through these layers thus establishing a low Richardson number zone under a more stable layer with the tropopause structures. Such an environment could favor shear-induced wave breaking phenomena in between the two tropopause layers, as Kelvin-Helmoltz instability would have been possible along the interface between the stably stratified layer and the unstable layer near 100 mb. Additionally, strong buoyancy-driven vertical motions associated with local deep convection could have lead to parcels being lifted into the stable region, producing a scenario where negatively buoyant downdrafts became favorable for the downward transport of southwesterly momentum aloft into a more northerly flow regime below. The result could be the focusing of rotors and significant turbulence kinetic energy near the incident location.
Session 11, Turbulence and Wind Shear II
Thursday, 21 January 2010, 3:30 PM-5:00 PM, B314
Next paper