The simple view of the tropical cyclone boundary layer winds would be of a steady decrease from gradient flow at the top, towards something variously estimated at 0.7 to 0.9 times that, at 10 m, together with a progressive turning of the wind direction towards the storm centre. Although the magnitude of the wind reduction factor is well supported by the data, there is a growing body of evidence to contradict the idea of a monotonic profile.
Indeed, a significant proportion of the few observations of wind profiles in the tropical cyclone boundary layer show a low-level wind speed maximum. The advent of a new observing technology, the GPS dropsonde, for the most recent Northern Hemisphere hurricane season, has heightened interest as it appears that the majority of soundings show such a maximum.
In this paper, we present a physical mechanism to explain the formation of the jet as due to strong inward advection of angular momentum. We then use a numerical model to verify this, and to show that the necessary inflow is maintained in spite of the presence of supergradient winds by vertical advection and diffusion. The numerical model also allows us to show that the jet is preferentially located on the left hand side, or opposite to the strongest winds, in a moving cyclone.
Finally, we show that the multiplying factor commonly used to reduce flight level winds, or gradient winds from parametric models, to the near-surface is affected by this mechanism and should have some spatial variability. In extreme cases it may exceed 0.9, although fortunately the larger values occur on the weaker side of the storm. It may, however, help to explain strong "second wind" damage as was observed, for instance, in Hurricane Andrew.