UPDATE: Thanks to a reader who pointed out where to find the 2008 version of BOM rainfall data, I’m able to plot the two data sets. There are differences. See addendum below. – Anthony
BOM loses rainfall
by Tom Quirk on Quadrant Online
Shock Murray-Darling Basin discovery
Analysts at the Australian Bureau of Meteorology have some explaining to do.
In the last two years some 900 mm of rainfall have been removed from the rainfall record of the Murray-Darling Basin. This startling discovery was made by comparing the annual Murray-Darling Basin rainfall reported on the Bureau of Meteorology website in August 2008 and the same report found yesterday.
The annual rainfall figures are shown as reported in October 2010:
Yearly rainfall in the Murray-Darling Basin from 1900 to 2009 as reported in October 2010 with a mean value of 467 mm (solid line).
There is no significant trend in rainfall through this period but there is large variability with rainfall extremes of a 257 mm minimum and a 787 mm maximum.
The comparison with the August 2008 report is revealing. The difference is a decrease of 900 mm rainfall in the 2010 report. The significant decrease occurs after 1948:
Changes to Bureau of Meteorology record of Murray-Darling Basin rainfall. Data downloaded August 2008 and October 2010
The Bureau is already on record adjusting Australian temperature measurements and they now appear to have turned to rainfall, making the last 60 years drier than previously reported.
One can understand that adjustments might be made to a few of the most recent years as records are brought up to date but a delay of forty or fifty years seems a little long.
This raises the question how certain is the data that is used by policy makers?
When we are confronted by apparently definitive forecasts of our future with rising temperatures and less rain, are we living through a period that brings to mind the Polish radio announcement of Soviet times?:
The future is certain only the past is unpredictable.
============================================
WUWT Reader Charlie A writes:
The Wayback machine comes to the rescue!
http://web.archive.org/web/*/http://www.bom.gov.au/web01/ncc/www/cli_chg/timeseries/rain/0112/mdb/latest.txt is the Murray Darling basin annual rainfall archived by archive.org on January 30th, 2008.
The current series can be found at http://www.bom.gov.au/web01/ncc/www/cli_chg/timeseries/rain/0112/mdb/latest.txt
The query form for data and graphs is at http://www.bom.gov.au/cgi-bin/climate/change/timeseries.cgi?graph=rain&area=mdb&season=0112&ave_yr=0
So, now having the older data, I decided to run a quick plot to see the difference. I fired up my Dplot program and came up with this in about a minute:
There are differences in the two data sets. The 2010 data has lower values, starting around 1950, just like the Quadrant article. WUWT?
– Anthony
Addendum: I notice the peaks post 1950 seem to be reduced, but the lows are not. Strange.
Discover more from Watts Up With That?
Subscribe to get the latest posts sent to your email.
Mike says:
October 15, 2010 at 6:54 am
Tim F wrote: “Before speculating or accusing further, it might be prudent to actually get some facts. That is, after all, how science (and reporting for that matter) is supposed to be done.”
Tim, people don’t come to WUWT for facts. They come for confirmation bias. http://www.skepdic.com/confirmbias.html
Got’a keep the costumer satisfied.
******************************
Mike – there are no facts in the experimental and natural sciences, there are only observations. This article shows 2 differing sets of observations. The difference is unexplained, but it may be that as a result of the article a reader (or the BoM) might point us in the direction of possible explanation. A main reason I visit WUWT is to read this sort of discussion, especially as I live in a country, like Australia, where the officials (NIWA) are known to make significant adjustments to the historical temperature record, leaving no audit trail or explanation for their adjustments. They also rely on a decades old study by a PhD student, which has not been peer-reviewed. My judgment is not easily swayed, I visit WUWT because I like open discussion of these and other issues, which is not easy to find on the internet.
Why do you visit WUWT?
All the best.
Person of Choler says:
How about this: When making adjustments to previously published data it behooves those making the changes to announce the adjustments and explain what they did and why.
That is, after all, how science is supposed to be done.
Well, what do you know. That’s what they did!
“Mean rainfall timeseries are calculated using a high-resolution gridded dataset developed for the Australian Water Availability Project (AWAP). The dataset uses topography-resolving analysis methods applied to all available monthly rainfall data passed by a series of internal quality tests. The resulting analyses have a resolution of approximately 5km x 5km and represent a substantial improvement on the grids previously used for operational rainfall monitoring.”
http://www.bom.gov.au/climate/change/about/rain_timeseries.shtml
So in conclusion:
* they developed a better way to analyze the data
* they presented the better data
* they told us that they had presented better data.
Between the explanation Tim F found and Charlie A’s analysis of the difference in trends between the datasets, it seems to me like there really is nothing significant to look at here unless somebody starts trying to make some kind of unsupportable point from it about
global warmingclimate changeclimate disruption.REPLY: The central question is: Why do rainfall records need to be adjusted at all? -Anthony
Tim F
October 15, 2010 at 11:53 am
” The dataset uses topography-resolving analysis methods applied to all available monthly rainfall data passed by a series of internal quality tests.”
So in conclusion:
* they developed a better way to analyze the data
* they presented the better data
* they told us that they had presented better data.
#
Cool. Probably one of the internal quality tests is to reject monthly readings that are outside a given threshold. So have we seen this sort of filtering before that just happens to have only rejected readings on the contra-narrative side?
Anthony, regarding your Update, you should be able to use the Wayback machine to get at the 2008 data for the other regions and for Australia as a whole. Were they all adjusted downward, or was the Murray Darling Basin a fluke? In particular, how does the current total for Australia compare to the 2008 data?
Mike S. says:
October 15, 2010 at 12:19 pm
Between the explanation Tim F found and Charlie A’s analysis of the difference in trends between the datasets, it seems to me like there really is nothing significant to look at here unless somebody starts trying to make some kind of unsupportable point from it about global warming climate change climate disruption.
Too late! There have already been dozens of unsupportable points made. The motivations and abilities of good scientists were assumed to be flawed by people making unsupportable points because of anti
global warming climatechange climate-climate disruption ideology.“The central question is: Why do rainfall records need to be adjusted at all? -Anthony”
To me, the central question is “Why were so many people ready to believe an unchecked, unvetted opinion from an unknown author?”
Across Australia generally, the second half of the 1900’s were wetter than the first half and that is reflected in the Murray-Darling Basin records. Although official records have not been reconciled for the 1800’s, there are sufficient indications that the 1800’s were drier and more drought prone then the 1900’s which indicates, that despite the latest drought that may have been etched in peoples TV memories, the trend is not as dire as many would have us believe.
Anthony asks the faux naif question
But the answer is clear. They do not fit with the theory. And we simply must not have theory and experiment disagreeing – it would be entirely unscientific.
So a few minutes work to make corrections to some old unimportant records and voila everything is back in order once more.
We can all sleep easy in our beds again. AGW re-confirmed..now with even more evidence of its correctness!
After a day of continuous rain yesterday, there is intermitment snow here in Canberra this morning. Occasional morning frosts in October are normal, but not snow!
As far as I am concerned rainfall simply … IS.
If more stations have been added, if conditions in hitherto unmetered areas are such that an overall change is noted why is this not given along with the raw data?
The “alteration” of rainfall figures cannot, under any circumstances, be tolerated.
If you get the “new improved” data and put it alongside the old data and plot the difference year by year you observe 3 stages of data improvement.
From 1900 to 1910 the new data averages 6.3 mm extra rainfall per year. With zero negative adjustments over that period.
From 1911 to 1945 the new data averages 3.4 mm less rainfall per year, with only 3 out of 35 readings showing a positive adjustment. This gives the combined average for 1900 to 1945 only 1.3 mm less rainfall per year.
From 1946 to 2007 the new data averages 14.1 mm less rainfall per year than the old data. There is only one positive correction out of the 62 readings.
So from 1911 to 2007, of the 97 annual readings there are only 4 positive adjustments.
I would love to hear the BOM’s rational for the changes but I doubt they will see the light of day.
The time series graphs from BOM has a toggle above that allows a trend line to be drawn like here.
http://www.bom.gov.au/cgi-bin/climate/change/timeseries.cgi?graph=rain&area=mdb&season=0112&ave_yr=T
Anthony writes:
The only reason I can think of is to…dare I use the word?…homogenize the data over a wider area. A rain gauge measures rainfall at a certain location over a certain period of time. Maybe some genius is trying to extrapolate the history at a single location over a wider area and a longer time.
I’ve always wondered about those weather forecasts that estimate a “40% chance of rain” in my area. Do that mean that at my location there will be a 40% chance of rain over the next 24 hours, or 100% chance of rain during the next 24 hours over 40% of the forecast area? Or maybe they do a root-mean-square calculation or something?
Sometimes I sits and thinks, and sometimes I just sits…
The Queensland Government used to produce a chart of Australia’s variable rainfall from 1890 – discontinued in 2004 but still to be seen on this site. It would be great if a private company could reproduce them again because it is very instructive looking at the chart because you can see clearly that while droughts have occurred, the trend in Australia appears to have been for more rain over the 20th century, not less. Each home in Australia should have one of these on the wall!!
Tim F says:
October 15, 2010 at 1:17 pm
To me, the central question is “Why were so many people ready to believe an unchecked, unvetted opinion from an unknown author?”
Generally, if data from the past is adjusted, there are 2 big reasons why:
1. purely scientific – better methods than were used in the current system and to get the most scientific answer the best methods must be used.
2. purely political – the old answer goes against political will so someone is tasked with producing something that better fits with what the politicians want to be done (politics can be governement of corporate).
Commenters have posted that there is political interest over the area so it is at least plausable that something else is going on. Is there no point in wondering exactly why/how it was changed? Or is it just no point in posting it on a blog for others to read?
oops… (politics can be governement OR corporate)
More on the non-existent decline in NSW rainfall
http://www.warwickhughes.com/sydwater/sydrain.htm
One of my first posts at RC years ago was countering someone from Australia railing on-and-on about the decline of rainfall on the continent in the 20th century, particularly in her area. Her proof was a BoM article with comments to the effect. The funny thing was that you could go to the BoM site and both download the data and have the website generate graphs for the continent as a whole or specific regions…and the results showed the trend was basically flat.
The second graph looks more like a step change at about 1950 or so. Equipment change? location change ?
As we shiver through a cold / wet October day in south-eastern Australia, with major flooding in numerous parts of the Murray / Darling basin, it is interesting to look at today’s (16/10/10) rainfall figures for the Canberra area and note how variable they are. Canberra airport (new station) 15.8mm, Canberra airport (old station) 17.0mm. These are 500 metres apart. Tuggeranong 48.8 mm (13km S of the airport) and Mt Ginini 73.4mm (41km SW of the airport). And our state of the art plastic guage on the back deck (5km SE of the airport) 33mm.
Try coming up with a ‘high resolution 5km x 5km gridded dataset’ with figures like this.
Rainfall can vary widely, as discussions between neighbouring farmers often shows, 😉 however over a period of time things do tend to average out, over a year as well as over decades. There are still however areas that are in rain shadows or the like that consistently are either higher or lower than surrounding areas
Tim F says:
October 15, 2010 at 1:17 pm
……..
“The central question is: Why do rainfall records need to be adjusted at all? -Anthony”
To me, the central question is “Why were so many people ready to believe an unchecked, unvetted opinion from an unknown author?”
Thats a very good question Tim F. Why don’t you actually go away and think about it. What is it that has caused such a massive amount of distrust in the data presented by climate scientists.
If what was being presented was a sound and incontrevertible as the AGW movement has been trying to tell us, there is absolutely no way there should be the level of mistrust and scepticism in climate “data” that we see today.
Just as trust is earned so is mistrust.
Ooops typo. Incontrevertible should be incontrovertible.
Data can dissapear but the rain won’t. I have a friend who just emigrated to Queensland from Scotland. He says he feels right at home.
DavidS says:
October 15, 2010 at 5:08 pm
Data can dissapear but the rain won’t. I have a friend who just emigrated to Queensland from Scotland. He says he feels right at home.
===================================================
Would have to agree, only 17.1c at the Hervey Bay Airport at 10:30 AM and where I am, 100meters from the beach, it is only 16.5c.
This time of the day here in the tropics, it should be around the 27 / 28c.