The purpose of this paper is to explore the feasibility of generating accurate mean vector wind statistics for every hour of the day from an hourly subset of cross-tabulated direction and speed data, using a 20-year test data set of hourly wind observations from the
Pt. Mugu, CA weather station. Twenty-year climatological mean vector
wind statistics will be generated for every hour of the day and month
(36-point, nearest knot observations) for the period 1973-1992, and these figures will be compared with results produced from a "synthetic" every-third-hour 16-point-Compass/Beaufort Scale scheme using the same original data. Estimated mean vector wind statistics for the "every-third-hour" selection (0100LST, 0400LST,...,2200LST) are derived from joint mid-value category magnitudes applied to the original observations, those for the "off-hours" from cubic-spline interpolation using the every-third-hour estimates of the u and v-components. Results are interpreted statistically and graphically, the latter employing 1) diurnal hodographs for selected calendar months and 2) a chart that plots mean vector wind arrows for every hour and calendar month on a single chart, overlain by constancy contours. This approach, if considered applicable on a wider basis, opens up the possibility for the quick, efficient, and accurate creation of station-by-station products of these kind from ISMCS-like data sets.