BAMOS Vol 32 No.2 June 2019 | Page 11

BAMOS Jun 2019 Figure 3: Scatter diagram showing direction and speed changes for the cases of abrupt wind speed increase. The change in direction is simply taken as the difference between the initial and final compass bearing over the half hour. Positive (clockwise) changes are referred to as veering while negative (anticlockwise) changes are backing. However, the change in direction is limited to 180 degrees, leading to ambiguity if the directional change exceeded 180 degrees. Examination of the data shows that such cases are rare. It should be emphasised that the discussion here concerns the north of Port Phillip which is a large body with many variations in wind across it. In particular, there may be large variations in wind strength from north to south in strong easterly or westerly flow. Overall Results There were 248 cases of abrupt increases in wind speed of 16 knots or more over the 21 years of data representing about 12 cases per year. There were 56 cases of abrupt wind speed decreases. Wind Speed Increases Figure 3 shows a scatter plot of the speed and direction changes. One third of the events (86 cases out of the 248) have direction changes less than or equal to 30°, and in the discussion below are termed “straight line” wind speed increases. For the cases with a change in wind direction more than 30°, two-thirds (108 out of 162) were cases of a backing wind and one third associated with a veering wind. Of note is that in four events the wind speed increased by more than 30 knots within 30 minutes. All were associated with significant backing in wind direction. Two examples are shown in Figure 4. Around 4am on18 November 2001 winds increased from 7 to 39 knots between successive 10‐minute data points. Winds stayed around 30 knots for most of the remainder of the day. An even sharper but shorter-lived event was 14 March 2009. Around 6:30pm the wind speed increased from 8 to 47 knots in a 20 minute period. The synoptic situation in each case involved a low moving over Victoria. In nine cases there were abrupt increases and decreases within adjacent or overlapping 30 minute periods. A review of available weather radar data showed that these events were typically associated with a line of thunderstorms crossing Port Phillip. Figure 4. Wind speed and direction at Fawkner Beacon for (a) 18 November 2001 and (b) 14 March 2009. 11