The Probability Distribution of Sea Surface Wind Speeds

Air-sea exchanges of momentum, energy, and material substances are strongly influenced by sea surface wind speeds. Because exchange fluxes are generally nonlinear in wind speed, mean fluxes are determined not just by the mean wind speed but also by higher order moments. This fact motivates the development of models of the probability density function (pdf) of sea surface wind speeds. This talk will present characterisations of the pdf of sea surface wind speeds on a global scale obtained using SeaWinds scatterometer observations. It will be shown that the skewness field of wind speed is characterised by three distinct provinces: the Northern Hemisphere midlatitudes, the tropics, and the Southern Ocean. These provinces reflect the fact that the skewness is a decreasing function of the ratio of mean wind speed to the standard deviation, such that the skewness is positive for small values of the ratio, near zero for intermediate values, and negative for large values. This relationship between moments is characteristic of the Weibull distribution, which has been used in the past to model surface wind speed pdfs. However, distinct non-Weibull structure is evident in the observed sea surface wind speeds. A mechanistic model for the pdf of sea surface wind speeds, constructed from basic boundary layer physics, is used to demonstrate that the observed relationship between wind speed moments is a consequence of non-Gaussianity in the sea surface vector winds which follows from the nonlinear dependence of surface drag on surface wind speed. This model characterises the observed sea surface wind speed pdfs at least as well as the Weibull distribution and has the benefit that it mechanistic rather than empirical.