Observed and simulated modes of vertical wind shear variability over the tropical Atlantic

Vertical wind shear plays an important role in modulating hurricane variability in the tropical Atlantic region. The ability of global climate models to simulate the statistical properties of observed vertical shear is analyzed, using the suite of model integrations performed for the IPCC-AR4 assessment. Significant biases are found in the simulation of the mean shear during the hurricane season, especially over the Main Development Region (MDR). Statistical properties of vertical shear are analyzed from daily to multi-decadal timescales. On sub-seasonal timescales, vertical shear is shown to have significantly non-gaussian properties. On longer timescales, the variability of vertical shear is analyzed using Empirical Orthogonal Function (EOF) and Singular Value Decomposition (SVD). A dominant dipole mode of vertical wind shear is found at both interannual and multidecadal time scales. At interannual timescales, this dipole mode is related to the remote SST gradient between East-Pacific, West Indian Ocean and Indian-Pacific warm pool. At multidecadal time scales, the dipole mode is related to local SST forcing from Atlantic Ocean as well as remote SST gradients. Uncoupled atmospheric model simulations are also carried out to elucidate the mechanisms behind the dominant mode of shear variability.