Hear the words of Dylan

You don’t need a weatherman

To know which way the wind blows

When a link to this was posted on the Mallacoota Community Facebook group the person posting noted that "The answer is blowing in the wind."!

I started to think about this post when trying to come up with some way of comparing the readings from my weather station with the readings from the BoM site about 5 km away.  

Before getting to that I will put in a few words about quality.  Accuracy is one element of quality, indicating how close a measured data item is to the true value (where that can be detrmined).  To some extent how close you get is determined by how much effort you can put in (ie cost) and how long you want to take to do it.  So time and resources are also elements of quality.  These can be brought together though a sense of fitness for purpose: brain surgery requires a different level of accuracy to chain-sawing firewood.  For my purposes I am interested in broad level of accuracy suitable for informing general interest discussions, not launching space shuttles (nor even general aviation) so I can accept some level of rubberiness.

As I have commented in the past my weather station is not ideally situated being about 2m above the ground and rather sheltered from the East and South.  As would be hoped the BoM site is in an exposed position with the anemometer 10m above the ground.  What I was hoping was that I could find some material on the BoM site to form a standard against which my data could be assessed.  The happy (or at least less unhappy) hunting ground for this seemed to be a page of climate statistics.

Wind Run

Of the various measures of wind available I was initially attracted to the mean daily wind run.  The attraction of this comes from having just read "The Home of the Blizzard" by Mawson.  It is the story of his Antarctic adventures and he and his companions usually recorded the wind run as well as any outstandingly bad gusts.

From looking at the data from my weather station for March and April 2019 the run was approximately half the average long term value at the BoM site.  (Wind run is not in the daily or monthly climate statistics published by the BoM.)  Interestingly, the value at the BoM site for Mallacoota was approximately half that for the very exposed Gabo Island site (16 kilometres away).

The data sets are for different periods, with that for Gabo commencing in 1957  as opposed to 2003 for Mallacoota.  However I am doubtful that this would explain much of the difference.  So the difference between my site and the BoM could simply reflect a real difference in micro-climate at my site.

The mean daily wind run recorded at my site in April was 88.4% of the value recorded in March.  For the BoM long term mean, April is 89.0% of the March value.  I thus conclude that on this measure my weather station shows an expected level of change between the months.

Before leaving the subtopic I'd note that dividing the wind run by 24 gives the average wind speed across the day. Thus, for April the average speed at my station was 5.08 kph.  The chart below illustrates the pattern of winds through the day in April.

The black line is the average across all days in April  while the blue line is the day with the strongest wind (ie highest gust reading) and the red line is the day with the lowest gust.  So averaging out at ~5kph is correct.

My conclusions from this are that 

• Wind run is an appropriate measure to use as an indicator of windiness,as it includes a lot of information; and
• The information from my Weather Station is a reasonable measure of relative windiness for Mallacoota.

Recorded speeds

In the usual records for a day the BoM shows three measures of the wind:

• The highest gust in a day; and 
• the actual wind speed at 0900 and 1500.

Since BoM doesn't do daily wind run I have decided to choose one of these three measures as my indicator of current windiness.  My choice is driven by the way in which the long term monthly averages for each of them correlate with the long term mean daily wind run.  

To my surprise there is a negative correlation between maximum monthly gust and wind run (R= -0.34).  The other two series show a very significant positive correlation.  For the 0900 value R = 0.89 while for 1500 R = 0.92.  Here is a graph showing the BoM daily means of Wind Run and 1500 Wind Speed.  NOTE that the two measures are plotted on different vertical axes which emphasises the similarity in pattern across a year.

In addition to having the higher correlation the 1500 speed occurs after the day has developed a reasonable weather pattern  so I'll choose that as my standard indicator of current windiness from now on. 

My final chart shows a comparison of 1500 Hr wind speeds for the BoM Mallacoota site (BOM) and my Weather Station (WS) for the months of March and April 2019.  I have included some polynomial trend lines to filter out the random variation (aka statistical noise): as indicated by the pathetic values of R² the is a lot of noise, but I think it shows the similarity of pattern!

The chart also demonstrates that while the pattern is similar the values from WS are well below the BoM readings.

Wind Direction

Following  from the surprisingly good outcome from comparing wind speeds I thought I would take a look at wind direction.  In so doing I was not feeling very confident due to my weather station seeming to be fairly sheltered from the SW - E.  However I decided to extract the direction given by BoM for the wind readings at 0900 and 1500 for each day of April and compare them with the equivalent readings from my weather station.

I had a "duhhh" moment when initially attempting to do this since in some cases the BoM had a value of NNW (=338^o) and the WS recorded N (=0^o) or NNE (=22.5^o).  Although they are quite close compass points there is a huge difference in degrees.  Thus there was a poor level of correlation.

That was solved by manually calculating the differences as summarised in this chart.

In 44 (73%) of cases the difference was within one compass segment.  To my mind that is an astonishing level of consistency and indicates that the wind direction data from my weather station is also fit for the purpose of indicating, as Mr Zimmermann warbled, "which way the wind blows".