Recorded presentation2B.1
A new set of mean soundings for the tropical North Atlantic and Caribbean Sea
Jason P. Dunion, NOAA/AOML/HRD, Miami, FL
The Jordan mean tropical sounding has provided a benchmark for representing the climatology of the tropical North Atlantic and Caribbean Sea since 1958. However, recent studies of the Saharan Air Layer (SAL) have suggested that the tropical atmosphere in these oceanic regions may contain two distinct soundings (SAL and non-SAL) with differing thermodynamic and kinematic structures and that a single mean sounding like Jordan's does not effectively represent these differences. This work addresses this possibility by using GOES (split window) SAL imagery to examine over 6,000 rawinsondes from the tropical North Atlantic Ocean and Caribbean Sea during the 1995-2002 hurricane seasons. It was found that a two-peak bi-modal moisture distribution (dry SAL and moist non-SAL) exists in this region and that the Jordan sounding does not represent either distribution particularly well. Additionally, SAL soundings exhibited higher values of geopotential height, unique temperature profiles, stronger winds (with an enhanced easterly component) and higher atmospheric stability compared to the moist tropical non-SAL soundings. Given the extent of these differences and the relatively large occurrence of SAL soundings in the 1995-2002 dataset, the Jordan mean tropical sounding may need to be updated to provide a more robust depiction of the thermodynamics and kinematics that exist over the tropical North Atlantic and Caribbean Sea during the hurricane season. The results presented here could have important implications related to our understanding of the climatology for this part of the world and whether or not the use (e.g. in climate and modeling studies) of the Jordan “hurricane season” sounding as a standard background state for this region is appropriate. Recorded presentation
Session 2B, Tropical Climatology and Climate Change
Monday, 28 April 2008, 10:15 AM-12:00 PM, Palms E
Next paper