April 2019 Weather report

This post will be a summary of weather observed in Mallacoota in this month.  As I only have about 10 weeks of data, the longer term information will be mainly drawn from the BoM site, located off Betka Rd near the Airport.  I hope it will be clear which site the data is referring to, but they generally show a broadly similar pattern.

Overall Summary

The month of April was quite warm with average maximum and minimum temperatures both above average.  Its thus no surprise that the mean temperature was 1.04^oC above the long term average for the month.

The month was also quite dry with rainfall well below average as was Relative Humidity at both 0900 and 1500 hrs.

There was a particularly windy period at the end of the month.

Rainfall

There have been quite a few comments in the media about Adelaide and Melbourne having the lowest rainfall for the period January 1 to April since records begin (180 and 160 years ago respectively).  It wasn't that bad in Mallacoota: we have records covering most years back to 1975 and 6 of those years are below the 177 mm recorded in 2019.  

1981 looked to be a dry year but there was no rainfall observation for February so I couldn't us that year.  To get an estimate of what went on I looked at the BoM records for Gabo Island.  That series shows a good fall in February 1981 so it would not have been a low year at Mallacoota.  (As an aside, Gabo has a solid record fr rain back to 1867, so there is lots of information to be mined there!)

For the month of April BoM hs recorded 27.8 mm while my Davis WS recorded 30,8 mm.  The daily totals vary considerably between the two series.  In part this is due to the timing of recording with BoM records ending at 9am and my Davis recording on a day ending at midnight.  For this month however the major difference was a storm cell on 24 April which dropped 14.8 mm on my WS but only 2 mm at the BoM site,  (A number of other residents commented that they received falls around 15 mm from the event,)

The  following chart shows a comparison of BoM data for 2019, 2018 and the average since 1985.

 Clearly the month was well below average (but a lot damper than 2018!

I have used the daily falls - from my WS - and the average fall on that date from BoM to estimate the total annual fall.  By 30 April the projected total fall for 2019 was 720 mm.  The movement of the estimate through the year to data is illustrated in this chart.

Temperatures

The issue of timing is crucial.  The BoM approaches (note plural) to this are summarised here.  On a 'normal' day the maximum measured for a day ending at 9:am will be the "Daytime High Temperature" for the previous calendar day.  In contrast the minimum to 9am will usually have occurred early in the morning and thus match the "Overnight Low Temperature".  In situations where fronts come through (increasingly abnormal) there can be wide discrepancies!

My weather station readings are quite simple as they are based on a day ending at midnight, Australian Eastern Standard Time.  Note: I don't change the WS clock for daylight saving!

That being said the comparative averages for April 2019 are:

• Maximum: WS 21.8 BoM 21.0
• Minimum:  WS 13.4 BoM 13.3
• Average:    WS 16.9 BoM 17.1

Given the different physical situation of the two stations such similarity is remarkable.  The three values, for my WS are shown in the next graph.

Maximum Temperatures

The entries in this section are all based on the BoM site.  First a comparison of maximum temperatures for this year, last year and the average.

April 2019 was a fair amount cooler than 2018, but still above average.  The illustrate the data underlying the average this chart shows the average April maxima x year.

I have left the trend line in, but primarily so that I can comment that the value of R² shows that the trend is not significant.

Minimum temperatures

Again the entries in this section are all based on the BoM site.  First a comparison of minimum temperatures for this year, last year and the average.

 For minima 2019 was warmer than 2018 and the average.  The two year comparison probably reflects the more rain this year than last

While the longer time series looks to have an upwards trend again the value of R² shows that the trend is not significant.

Average temperatures

The BoM assesses average temperatures as the mean of daily maximum and minimum.  While not precise, I have looked at this in the past for my weather station and found it to be a very close approximation to the average of hourly values through a day.  

The following chart is rather (perhaps 'excessively' is the better word) complex but I hope illustrates a few things. An explanation of what's going on is below the graph.

The green line is the average of average temperatures temperature for that day,  There is a trend line associated with it, showing a significant downwards slope.  In simple terms the end of the month is significantly colder than the start!  While obvious, that wasn't the case in March.
The solid blue line is the average temperature month to date .  That is the average of the average temperatures from the first of the month to the date shown.  It also has a downwards slope, more significant that the green line, as the day to day variability has been averaged away.
The red line is the daily average temperatures calculated from this year's BoM extremes.  Much more variable.
The dashed mauve line is the averages from my weather station where each day has the actuall average of 24 hourly readings.  It is very similar to the red line, reinforcing the overall validity of the calculated averages.

Back to simplicity with a time series of averages calculated from the BoM data.

An important use of average temperatures is in calculating a temperature anomaly given by the difference between an average for this period and the average of all periods in the past.  In this application the anomaly is +1.04^oC.  That is the month was just over 1^oC above average.

A comment on global warming

For each month for each year I assessed whether the maximum or minimum were above the average maximum/minimum for that month.  I then counted the number of above average months for each year.

• For maxima the series of number of above average months per year, sloped upwatrds quite significantly (R²= 0.52).
• For minima the line slopes up but not significantly (R²= 0.26).
• Adding the 2 scores together shows an upwards slope with a close to significant R²= 0.49.

Although 25 years is quite a short period it does seem Mallacoota is warmer than it was.

Humidity

The BoM publishes 2 series for humidity covering the readings at 9am and 3pm.   The chart below shows the observations from my WS for those two times through April,  They show very similar patterns except for high 1500 readings on the 6th and 22 -24 which are leading up to rain.

The BoM have some average RH readings for the 2 times, but for some unknown reason they terminate at 2010. However that is good enough to give a context for current values.  For both times 2019 appears to be quite an amount drier than in the past.

Wind

This is a difficult topic as there are a number of ways of measuring wind (day-run, gusts, point of time etc) and my WS is not in an ideal situation.  The latter point is particularly important when compared with the BoM site which has the anemometer at 10m high, in an open site.

My starting point is to look at the maximum daily gust from my WS as illustrating the windy days during the month.

I have also looked at the daily windrun and the speed at 9AM (as both of these are available to some extent from BoM records).  The correlations between the 3 series from my WS are all quite good, at ~0.70. 

However when looking at point of time windspeed from the BoM site for 0900 and 1500, it can be quite different to that from my WS.  For the 0900 series the differences can be quite large (eg for one reading the WS speed is only 18% of that given by BOM).  While the correlation is again quite good at ~0.7 the differences are uncomfortable.  Further the differences vary between the 0900 and 1500 with the difference having a much lower correlation between the two series.

My conclusion from all of this is that my data is best restricted to showing when the windy periods occurred in the month using the chart above.