Jennifer Marohasy
I have been assured for the past few years, including by Australian Bureau of Meteorology Director Andrew Johnson, that the change from mercury thermometers to platinum resistance probes is Are not the cause of, nor a contributor to, global warming as reported on the nightly television news.
If so, this will become evident as the number of hot days and their average temperatures increase – much like what we are known to be caused by increasing levels of carbon dioxide in the atmosphere.
The simplest way to know the impact of the change on the probe – and to distinguish this from the potential impact of warming from carbon dioxide – is to compare the automatic reading from the probe with the result. Manual reading from mercury thermometers at many weather stations. many years.
The office collected this data as handwritten records on A8 forms. There’s no official list, but when I put the information together, I’m confident that parallel data – probe measurements against mercury – exists for 38 weather stations and from many of these stations. 20+ years of daily data will be available to allow comparisons. Access to all this information and its analysis will allow to assess some of the consequences of changing the device. The problem is doubly complicated by the office using more than one type of probe, varying the type of probe used and the type of data being transmitted electronically – averaging first and then moving to recording. immediate treatment. I detail some of my initial concerns in a letter to chief scientist Alan Finkel back in 2014, almost ten years ago.
More recently, I pondered this issue with Chris Gillham – from http://www.waclimate.net. We discussed applying our method to understand extreme rainfall trends to understand temperature extremes, or at least how the office can inflate daily maximum temperatures through equipment change, and how to quantify this even if we don’t have access to the data. whether parallel.
While I was a bit distracted with the Administrative Appeals Court mediation, Chris Gillham had just started work on downloading the unadjusted temperature data and calculating the 99th percentile (the hottest 1%), 95th percentile (5% hottest), and 90th percentile (10% hottest) for all locations with automatic weather stations used by the Australian Bureau of Meteorology to calculate change and variability climate. These are locations known as ACORN-SAT (Australian Climate Observation Reference Network – Surface Air Temperature). There are 112 weather stations that make up this network, and 105 of them are automated with platinum resistance probes as the primary instrument of official temperature recording.
Chris Gillham downloaded the data and arranged it so that the transition to an automated weather station (AWS) start date with synchronized platinum resistance probes. For clarity for some locations, e.g. Bridgetown, AWS was installed in 1998 while in Birdsville, AWS was installed in 2001. For example, to calculate 1% of the hottest days annually starting starting 20 years before AWS was installed and then for many years after that as we do in a forthcoming report, the starting years need to be consistent.
It was a huge undertaking, with the data analysis done. All will be published as reports in due course, with results for many individual stations eventually available via spreadsheets stored at http://www.waclimate.net. Chris calculated three different categories (99th percentile, 95th percentile, and 90th percentile) based on all daily observations since 1910 (yes, a really huge undertaking).
The question is whether platinum resistance probes in automated weather stations increase the frequency of extreme temperature observations, most likely because they have faster response times than thermal mercury meter.
Let me summarize some of the findings here:
At most stations, the frequency and annual average temperature of the 90th (hot) and 95th percentile days (very hot) both increased rapidly when their first platinum resistance probes were installed, with the 99th percentile (extremely hot) showed an increase in frequency but little effect on mean temperature.
As the frequency of the maximum percentiles increases, the mean temperatures within those percentiles usually also increase.
There are inconsistent results between multiple stations. For example, at many stations, frequency increases sharply within the 1st, 5th, 10th, 90th, 95th, or 99th percentiles when AWS is installed, as well as the average temperature within those percentiles. However, at some stations, the frequency increased but there was no change or decrease in temperature in the percentile. In addition, the frequency can decrease while the temperature also decreases, stays the same or increases.
Likewise, a station may show an increase in the 90th percentile frequency or mean temperature, but a decrease in the 99th percentile frequency or mean temperature.
This could be an indication of different brands of platinum resistors being installed and/or that each probe of any brand has different response characteristics in terms of how frequency or temperature is recorded.
In addition, the AWS installation may involve another environmental variable, such as a slight change in location or the corresponding installation of a small Stevenson display that changes exposure to air. hot or extremely hot air in different ways.
Only through the general average of all synchronized observations can the effect of the AWS installation be established.
At most stations there is a rapid change, higher or lower, in frequency, average temperature, or both in the different extremes when AWS is installed, which shows platinum resistors do not record these extreme hot and cold temperatures in the same way as manual thermometers used to.
The majority of ACORN stations (69 vs 36) experienced an increase in frequency of the 90th, 95th, and 99th percentile days, with the collective average showing a sharp increase coinciding with AWS installation . Switching from manual readings from mercury thermometers to AWS observations causes frequency and temperature to increase in some stations and decrease in others, but the extreme temperature effects of AWS are more about heating than cooling.
The peak percentile is not stable after the initial spike, and there are periods of frequency and/or temperature increases in the years following the initial AWS installation. One possible reason for the peak values to increase gradually or sporadically is that of the 105 ACORN automatic weather stations, 59 have had their platinum resistance probes replaced at various times since the installation of AWS. initial.
Chris curated more data to look at the effect of alternative probes. Only six years are compared after probe replacement as this ensures all 59 stations are compared, with the most recent replacement being six years ago and including 2022.
Alternate probes at 59 ACORN stations have an effect on the frequency of peak percentiles and average temperatures, especially on extremely hot 99th percentile days. Considering the 1% of the hottest days after the replacement probe was installed, the temperature increased by 0.36 degrees Celsius on average, while these hottest days increased by 50%.
Considering Alice Springs, as an example of the effect of replacement probes on frequency and percent extremes, in the 5 years before and after the 2011 AWS probe replacement, Alice Springs increased by 27, 7% (44.8 vs 57.2 pa) of annual temperature frequency at or above the 90th percentile (10 years before and after 45.3 v 60.7 pa = 34.0% increase).
In the 5 years before and after replacing the AWS probe in 2011, Alice Springs experienced a 0.1°C increase in temperature at or above the 90th percentile (10 years before and after the 0.3°C rise).
In the 5 years before and after the 2011 AWS probe replacement, Alice Springs experienced a 23.3% increase (24.0 vs 29.6 pa) in annual temperature frequency at or above the 95th percentile ( 10 years before and after 23.5 compared to 33.6 pa = 43.0% increase).
In the 5 years before and after replacing the AWS probe in 2011, Alice Springs increased 0.3 degrees Celsius in temperature at or above the 95th percentile (10 years before and after 0.4 degrees Celsius).
In the 5 years before and after the 2011 AWS probe replacement, Alice Springs experienced a 48.0% increase (5.0 vs 7.4 pa) in annual temperature frequency at or above the 99th percentile ( 10 years before and after 4.4 compared to 11.7 pa = 165.9 % increase).
In the 5 years before and after replacing the AWS probe in 2011, Alice Springs experienced a 0.4 degree Celsius increase in temperature at or above the 99th percentile (10 years before and after 0.3 degrees Celsius).
It is worth noting that, for example, Alice Springs had only 76.8mm of precipitation in 2009, with 51 days at the 90th percentile (38.1C+), 23 days at the 95th percentile (39.6C+) and three days at the 99th percentile (41.8) C+). 2012 was much wetter 210.4mm so you would expect fewer hot, very hot and extremely hot days, but instead 74 days at the 90th percentile, 41 days at the 95th and 8th percentiles. days in the 99th percentile.
In summary, the 90th, 95th, and 99th percentile data currently available provide compelling evidence that platinum resistance probes in automated weather stations have increased the frequency of hot, very hot, and extremely hot days. in Australia since 1996, with further pattern change increasing since then installing replacement probes as recently as 2016.
Related
