Warmest winter on record for Victoria ? or BoM mistake ? (that’s Australian Bureau of Meteorology)

TV news and weather presenters are gloating lately as they report Australia’s “hottest winter ever”. I was traveling by car on the 27th August and Dr David Jones of the Bureau of Meteorology National Climate Centre was being interviewed by the ABC 666 just after 9am. I suppose “interview” is not quite correct, a mutual gush session might be more accurate. Dr Jones was talking up the notion of our “hottest winter” despite there being a few more days yet to run.
Anyway, the BoM now have a new media release which is slightly less trumpeting. They say we just missed the “hottest winter” label except for New South Wales, Victoria, and South Australia.
To get a handle on our winter warmth you can make contour maps of maximum temperature anomalies and minimum temperature anomalies at this useful BoM webpage.
Maximum temperature anomalies
and at night time
Minimum temperature anomalies
Remember from the above BoM media release that the state of Victoria is mentioned as actually experiencing its “warmest winter on record”.
Now check the state of Victoria on both maps I say Victoria has been too near average this winter to have a “snowball chance in hell” of having its “warmest winter on record”.
So I say to the BoM – if your maps are right or near right – then your claim that Victoria has just experienced its “warmest winter on record” has to be wrong.

12 thoughts on “Warmest winter on record for Victoria ? or BoM mistake ? (that’s Australian Bureau of Meteorology)”

  1. I’m a pure layman, so please filter accordingly. In my closest town, Kempsey, our driest winter months were in the late nineteenth, early twentieth centuries. The hottest August was 1946. 1918 was the hottest June, 1917 the hottest July. Similar but even older dates for hottest spring months. I realise that doesn’t make a single winter or spring aggregate, and that things like cloud-cover can make such statistics mean little; also, I may well be misreading the records. My concern is that our climate authorities seem to have even less sophistication than me. For example, today is a warmish, windy day such as you can often get in early spring, and lately the weather’s been unseasonably warm. Drought is now a worry.

    Yet this warm winter has something in common with the previous two severe winters: namely, winter/spring winds are coming somewhat less from the west. Even today’s nagging wind is not a classic “westerly” blowing from the interior. It’s almost a nor’easter. I remember travelling in Italy in ’05 and being surprised and disoriented by winter thunderstorms. Last year it happened right here!

    My question to those here with some expertise: is there something else operating here beside nino/nina. Is it back-to-the-fifties thanks to PDO? If my comments on wind-direction etc are the ramblings of an aging hobby farmer, please tell me so. But you’ll appreciate that nowadays it’s pointless to ask such questions of the established weather authorities.

  2. Nice to hear from you Jon. I can not make your link work, could you please email me that mean T map, thanks.
    Using station data at this site I have quickly found that 2005 had a warmer winter than did 2009 – for Melbourne Regional Office 86071, Rutherglen Research 82039, and Nhill Aero 78015.
    Great the way the SOI is still near ZERO too.

  3. Thanks Jon for the new link.
    The graphic indicates that 2009 was warmer than 2005 by only a small amount over Victoria. I maintain my claim that the BoM gridded data is not good enough to be sure that the BoM conclusion is correct. All you could say for certain is that 2005 and 2009 winters were comparably warm across Victoria.
    Note the SOI has come back near zero again – another BoM mistake claiming an El Nino this year.

  4. Jon and Warwick,

    I’m intrigued. Can’t work out exactly what is going on.

    Jon’s BoM graphic in link 4 shows Victoria’s average temperature 1.00 degrees higher than average for winter 2009. Warwick’s two graphs in his original post show maximum and minimum temperatures for winter 2009 in Victoria, and the way I read them, the minimum must be less than 1 degree above average (only about 20% of the state is over 1 degree warmer) and the maximum must be about 0.6 degrees above average (there is almost as much area below average as there is more than 1 degree warmer). Both minimum and maximum are clearly under 1 degree warmer than normal, so how come the mean is 1.00 degrees above normal? How can the average of two numbers under 1 be 1?

    In this deterministic universe of ours, there must be an explanation somewhere. The base period is the same (1961-90), so that’s not it. Is the mean not an average of maximum and minimum, but an average of hourly values? Has the BoM cocked up somewhere? Or is there some other reason?

    I am also unsure why the BoM graph only goes back to 1950. The BoM used to say it only trusted the data back to 1910. Sceptics wondered whether that was because temperatures appeared to be hotter in 1880. Whether they were right or not, 1950 is really pushing it. With 30-odd year oscillations in PDO and other climate patterns, data back to 1950 doesn’t even give us two switchovers, so we can’t compare 2008 with the period around 1945 that might be at roughly the same point on warm/cool cycles of this length.

    But my main doubt remains whether to trust the BoM figures at all. Older analyses of historical temperature data in Australia showed no big warming from 1980 as the current BoM analysis does. There has to be a lot of doubt about data from some of the BoM’s supposedly reliable climate monitoring sites in big cities. And they never seem to compare their data with those from satellites measuring the temperature a little higher up – why not?

  5. This is also something I have noticed. Just going on terribly un-scientific looking-out-your-window type of analysis, I questioned the “warmest winter” and went and had a little look-see at available data, which would call this claim into question. At the same time as this was reported, there were news stories of the extended length of the Victorian snow season. Something screwy going on here…

  6. Warwick – help.
    I have posted this under this article because I, like you, feel that the official figures don’t seem to reflect reality.
    I have just read the recently released climate summary for NSW for October 2009 at www.bom.gov.au/climate/current/month/nsw/summary.shtml.
    The summary says that NSW was25.2°C which is 0.7°C above the historical average of 24.4°C. I checked the anomalies max temp map and, like you, found that it did not reflect that conclusion. So I checked the summary statistics to see if I could check them against the reported 25.2C.
    There are 185 stations listed with the max means posted next to each. The next column shows the difference from the av max mean but only about half the stations have a result recorded in that column (I checked back and found that that this seems to be a random thing as some not recorded in Oct were recorded in other months). I noticed that there were more stations in what was the ‘cooler areas’ not included in the anomalies section than were missing in the ‘warmer areas’.
    So I added all the av max means and divided by 185 and found that the average mean for all stations was 22.8C, well below the figure of 25.2C.
    What criteria is used to determine the state’s average mean? Are all WS in the summary statistics used and, if not, why are some excluded? I understand that some stations may not be included if the temp data is incomplete etc but checking some of those in the summary statistics did not show this to be the case.

  7. Thanks for responding so quickly. I am just very curious as to what is happening here. I have double-checked my figures and I’m sure I am correct. I hope this is not a wild goose chase and there is a perfectly reasonable explanation – just something seems ‘fishy’.
    Maybe they only collate those that show up in the anomaly column.
    PS No need to reply to this – just musing.

  8. I have had a quick look at the BoM puff for NSW in October. They really are not claiming that much, it was a so-so month.
    But I agrre, their contour maps for monthly max T and min T have inconsistencies.
    Quickly doing a few checks on the monthly max T and min T contour maps, using the monthly totals you can get at the Australian Weather News site.
    Canberra looks in correct contour for max but wrong for min, according to Canberra AP should be in + o.5 contour.
    Armidale AP is wrong too, the BoM max contour map has it 1-2 degrees warm yet actual T was 0.5 cooler than norm, min T was OK.
    Deniliquin AP is in 1-2 deg contour zone on max map but actual T is only 0.5 above norm, on the Oct min anomaly map the 0 contour runs west of Deniliquin and the actual min T was 0.8 warm – so both max and min wrong there too.
    Notice on the max anomaly map the small brown bulls eye (+2 – +3 anomaly) ~between Wagga and Hay, looking at the high res map it might be Narrandera or Leeton. I can only find data for Narrandera Golf C, the actual max was 1.5 over norm, NOT 2-3 deg over norm, the min was 0.6 cool which looks OK.
    My eye was drawn to their red bullseye in WA (max anomaly map for Oct); take Laverton, which is in the 3-4 degrees + contour for max T and 2-3 deg + contour for min T. Laverton actual monthly max and min was +2.2 over norm and +1.8 over norm. So in this case their contouring “software” is exaggerating both max and min by a whole contour band.
    I think this is where the 0.7 statewide figure for the Oct max anomaly comes from Ian George. The figure that you originally concerned about when you first posted and believe is exaggerated.
    It is puzzling it is so easy to find discrepancies, their contouring must be very rough.

  9. I am still curious as to how the BOM calculate the av means for each weather station to arrive at their monthly state average.
    I added all the average max mean temps for each listed W/S in their NSW summary statistics and divided by 185 and found the mean below the state average, not the 0.7C above they state. I then only used the stations they listed in the anomalies column and found that was below average too. I then used the highest and lowest means for each district and that was below average as well. This is all at: www.bom.gov.au/climate/current/month/nsw/summary.shtml
    I have emailed the BOM but as yet have not received a reply. Does anyone know how the monthly means are calculated?

Leave a Reply

Your email address will not be published. Required fields are marked *