Open Mind

Australia 5+1

February 10, 2009 · 154 Comments

I recently posted about the heat wave in southern Australia. I emphasized that what’s really important is not the short-term events but the long-term trend, and showed that Melbourne, in addition to experience a brutally hot short-term event, also showed a long-term warming trend both in its mean and max temperatures. The data for Melbourne go back to 1855, and the strong increase in temperature was a recent phenomenon (especially for max temperature).


This led to the usual protests of “Urban Heat Island!” In response to that, one reader suggested that it would be a good idea to take a look at some not-so-urban stations in Victoria (the state of which Melbourne is the capital), and even suggested 5 stations which would be good candidates. I didn’t pick these stations, and by all appearances the one who did hadn’t seen the data. Let’s take a look.

For these stations we’re only looking for the trend, so it isn’t really necessary to use the daily data (there’s a lot of it!) we can do perfectly well with monthly averages. For each station I’ve computed anomalies to remove the annual cycle, and smoothed the result with a lowess smooth on an approximately 30-year time scale. First on the list (in alphabetical order) is Cape Otway:

capeotwa

Next is Deniliquin (which appears not actually to be in Victoria, but just over the line into New South Wales):

deniliqu

Then comes Mildura:

mildura

Next is Nhill:

nhill

And finally, Wilson’s Promontory:

wilsons

All five stations show a recent increase (trend, not event) in their max temperature. Four of the five show an increase in min temperature; only Nhill shows a decline. Of the 15 records (max, min, and mean for five stations), 14 show an increase over the span of observation (only the min temperature for Nhill shows a decline), although this should actually be counted as 9 of 10 because the mean is the average of max and min.

Some of the stations show very large warming; Mildura exhibits a 1.5 deg.C rise in max temperature over the recorded time (and a steady increase in min temp), Cape Otway a 1.4 deg.C rise in min temp, even Nhill a 1.3 deg.C increase in max temp. And at Wilson’s Promontory, min temp has increased about 1.8 deg.C since 1958.

Nine out of ten. Let’s make no mistake about it, the trend in Melbourne is not an isolated incident reflective of its sprawling urbanization — it’s the prevailing trend over the state of Victoria in Australia.

Another reader, known for his disbelieving tendencies, responded to the suggested list with this comment:


Craig, I have looked at the daily temps for Ballarat. The loess function does not behave anything like Melbourne’s.

I didn’t find any daily data for Ballarat among the high-quality data sets provided by the Australian BOM, or even any monthly data. But I did find annual averages. Since they’re annual, we don’t need to remove the annual cycle and can work directly with the data rather than anomly. For Melbourne I computed the pattern in mean and max temperature; here’s the same for Ballarat:

ballarat

How can you fail to see any similarity at all with the Melbourne record? Let’s have some of what you’re smokin’.

UPDATE
A few graphs

vic

vicanom

Categories: Global Warming
Tagged:

154 responses so far ↓

  • philscadden // February 10, 2009 at 11:31 pm

    There is a hint of a 20 year cycle in the Mildura/Melbourne annual averages. Indian Ocean Dipole? Present in the others if you dont have a 30 year smooth?

  • Craig Allen // February 10, 2009 at 11:47 pm

    Thanks Tamin. You’re a champion.

    And no I didn’t look at the data when suggesting the stations. I was just picking ones from the Bureau of Meteorology high quality reference set that I knew were least likely to be affected by a heat island effects. Cape Otway and Wilsons Promontory in particular are on isolated cliff-tops facing the Southern Ocean.

    The links to those stations again so readers can see the photos of the sites:

    Cape Otway #090015
    Wilson’s Promontory Lighthouse #085096
    Nhill #078031
    Mildura airport #076031
    Deniliquin Airport #074258

  • Hank Roberts // February 11, 2009 at 12:27 am

    http://bravenewclimate.files.wordpress.com/2009/02/heatwavemap.jpg?w=476&h=327

  • saltator // February 11, 2009 at 12:57 am

    Tamino,

    You obviously have a longer time series of data than I had access to. My data for Ballarat came from: http://www.mdbc.gov.au/subs/seaci/research_data.html which gives data between 1957 and end 2005.

    I did notice the increase toward the end of the series. However, the early part of the series (from 1960 on) the temperature anomaly fluctuated more than the almost constant increase in Melbourne’s temp.

    Why did you use temp for Ballarat and Temp Anomaly for the others?

    [Response: So you noticed the increase at the end of the series, but you still said "does not behave anything like Melbourne’s"?

    You probably aren't controlling the parameter (in R it's the "f=" parameter) which sets the characteristic time scale for the smoothing. With a shorter time series, if you keep the default paramater it will shorten the characteristic time scale, so you'll get more fluctuations for that reason alone (just as you get more fluctuations with moving averages when you use a shorter vs longer time scale).

    For Ballarat I had annual rather than monthly/daily data so there's no annual cycle. Hence the difference between temperature and anomaly is a constant, which makes no difference to the pattern of changes.]

  • Craig Allen // February 11, 2009 at 1:04 am

    The Bureau of Meteorology statement on the recent south-eastern Australian heat wave is here.

    It is remarkable how many places and regions broke temperature records by really substantial margins. And that this follows an unprecedented heat wave last summer that also broke lots of records.

    Where is the global cooling that all the denialists are promising us?

  • Stephen Spencer // February 11, 2009 at 1:57 am

    Thanks for another great post Tamino.

    At the moment the weather / climate topic of conversation is the recent heat wave, but the ongoing topic is how dry it has been in Victoria and southern Australia for the last decade.

    The Bureau of Meteorology (BOM) is now saying that the current “drought” is 12 years long. This fits with my perception of the situation watching the weather over the years. I have been using the BOM site to produce time series of precipitation around the Melbourne area and it looks to me that the current dry spell started in the mid 1980s.

    Would you be able to check precipitation in Victoria and confirm or deny my impressions?

  • saltator // February 11, 2009 at 2:52 am

    My R script for Ballarat is outlined below. Perhaps you could examine it for flaws in logic. I know that I use a lower f value than you. An f of 0.75 gives a similar transform to yours but I am advised that a high f of that magnitude is too coarse. I also used temp anomaly for maximum daytime temperatures.

    Script:

    url = “http://www.mdbc.gov.au/subs/seaci/datafiles/tmax/ballarat”
    bal = read.table(url, header=F, col.names=c(”day”, “dayYear”, “temp”), na.strings = “-999.9″)
    attach(bal)

    # convert row numbers to data
    dx = row(bal)

    # Assign x to dx column 1
    x = dx[, 1] # this gives a day number

    #Assign y to temp anomaly (subtract 14C from each daily temp as per GISS to give
    # the anomaly 14C = mean temp for base period 1951-1980)
    y = temp-14

    plot(x, y, type = “p”, pch=16, cex = 0.1, col = “red”, main = “Ballarat”, xlab = “Days from 01/01/1957″, ylab = “temp anomaly (max temperature)”)

    # Create Loess function
    y.loess = loess(y ~ x, data.frame(x=x, y=y), span = 0.30)
    y.predict = predict(y.loess, data.frame(x=x))

    # plot the loess transform over the main plot
    lines(x, y.predict, col = “blue”, lwd = 2)

    # Plot the loess transform in a separate window
    windows()
    plot(x, y.predict, col = “blue”, lwd = 2, type = “l”, main = “Ballarat loess”, xlab = “Days from 01/01/1957″)

    [Response: I generally use the "lowess" function in R rather than "loess," so there may be aspects I'm not aware of, but: if "span=0.3" means use 30% of the data per estimate (as it does for lowess), then with about 60 years of data you'll use about 18 years for each estimate. That's only a bit more than half of a 30-year span, so you'll get more fluctuations than a 30-year span.

    For the Melbourne data I used f=.25 on 155 years of data, so the time scale is nearly 40 years. For the other Australian stations I chose f so that the product fT (T=total time span of data) was 30 yr. But instead of using R, I used my own lowess program which has a slightly different weighting function, so results with R won't be precisely the same.]

  • saltator // February 11, 2009 at 4:11 am

    “if you keep the default paramater it will shorten the characteristic time scale, so you’ll get more fluctuations for that reason alone (just as you get more fluctuations with moving averages when you use a shorter vs longer time scale).”

    Actually, in R, the default is 0.75, therefore you woul get fewer fluctuations. I used loess instead of lowess as loess allows calculation with missing data. In loess in R f is called span.

    [Response: What I meant was that if you keep the parameter constant, then since the characteristic time scale is proportional to the product fT, shorter data sets (smaller T) will lead to smaller time scales and more fluctuation.]

  • saltator // February 11, 2009 at 4:50 am

    Ok. I understand where you are coming from now. Thanks for your explanations.

  • Tom G // February 11, 2009 at 6:09 am

    If Australia is our canary in the coal mine and I don’t doubt that it is…
    We are in trouble…serious trouble.

  • Polyaulax // February 11, 2009 at 6:54 am

    That decline in Nhill mean minimum anomaly is intriguing. Is there a decline in nocturnal cloud cover,and cloudy days in general?

  • Boris // February 11, 2009 at 1:04 pm

    Clearly these graphs prove that Australia had an increase in air conditioners, BBQ grills, parking lots and outdoor hair dryers starting in 1995.

  • luminous beauty // February 11, 2009 at 1:45 pm

    Steck,

    There’s no good reason to assume the 1951-1980 mean in Ballarat is the same as the global mean. ‘Specially since you are plotting daily max and not daily mean.

  • sod // February 11, 2009 at 2:00 pm

    but but but the UHI effect of the nearby parking space caused the temperature increase and the bush fires!!!

    ;)

  • paulm // February 11, 2009 at 4:55 pm

    Dear open mind,
    Do we know what caused the medieval warming?

  • Ray Ladbury // February 11, 2009 at 6:26 pm

    Paul M., Ever hear of Google?

    http://www.google.com/search?q=medieval+warm+period%2C+causes&rls=com.microsoft:en-us:IE-SearchBox&ie=UTF-8&oe=UTF-8&sourceid=ie7&rlz=1I7IBMA

    Do yourself a favor and stick to the references by actual climate scientists.

  • David B. Benson // February 11, 2009 at 9:52 pm

    paulm // February 11, 2009 at 4:55 pm — “Plows, Plagues and Petroleum” by W.F. Ruddiman is a good palce to start.

  • Neil Fisher // February 11, 2009 at 10:48 pm

    [Response: So you noticed the increase at the end of the series, but you still said "does not behave anything like Melbourne’s"? ...]

    A trend of less than 30 years - clearly, it’s weather noise and we can ignore it.

    Oh wait - that’s right, I forgot. That’s only true for cooling, isn’t it?

    [Response: Trends are decided by statistical significance, not length of time span. Trends from mainstream climate scientists consistently pass the test; those from denialists consistency fail.

    Oh wait -- denialists don't care about valid statistics, just about a snarky punch line.]

  • saltator // February 11, 2009 at 11:33 pm

    luminous,

    Just using the GISS baseline value to get an anomaly value. It should be ok but others are more qualified to verify that.

  • saltator // February 11, 2009 at 11:35 pm

    Tom G,

    Australia is not the canary in the coal mine. You have to live here to understand that.

  • Peter Martin // February 12, 2009 at 3:10 am

    It is interesting to look at the Australian BOM website for trends over the past 40 years for all land areas including Australia.

    http://www.bom.gov.au/cgi-bin/silo/reg/cli_chg/g_trendmaps.cgi?variable=global_t&region=global&season=0112&period=1970

    These show that the interior of Australia, like many other land areas far from the sea, has warmed by much more than the global average. To those deniers who may be unaware of the population density of this region, I’d just say that you’d be hard pushed to blame the measured increase on the UHI effect!

    Central Australia has warmed around 0.5 deg C per decade which would mean that these areas now are a couple of degrees hotter than 40 -50 years ago. The white areas on the map have insufficient data so we can’t say for sure exactly how big an area; but, when the weather systems from these regions move towards Melbourne or Adelaide in future years it is likely that recent records will themselves be broken sooner rather than later.

    Incidentally, I’d like to see more emphasis on the figures for temperature increases on land, where most of us live, rather than on the more usual land/sea average. I would suggest this tends to understate the seriousness of the AGW issue.

  • Peter Martin // February 12, 2009 at 3:24 am

    A better map of Australian average temperature changes:

    http://www.bom.gov.au/cgi-bin/silo/reg/cli_chg/trendmaps.cgi?variable=tmean&region=aus&season=0112&period=1970

  • Craig Allen // February 12, 2009 at 5:03 am

    Tamino,

    Following on from Philscadden’s question at the top about the Indian Ocean Dipole:

    The data is here

    I plotted it and there appears to be a change to the pattern since 1989. The fluctuation in the index are clearly lessened and it has stayed mostly positive.

    It this statistically significant? If not, how long would it have to continue before we could determine that it is is significant?

    And is there a significant correlation between the IOD index and southern Australian temperatures? (As represented by the Cape Otway or Wilson’s Promontory datasets say?)

    I feel a bit sheepish about suggesting work for you to do for us, but would be very interesting to know.

  • Lab Lemming // February 12, 2009 at 11:04 am

    “Tom G,

    Australia is not the canary in the coal mine. You have to live here to understand that.”

    Correct. Australia is the coal mine.

  • Chris // February 12, 2009 at 9:54 pm

    Re: Peter Martin: “Central Australia has warmed around 0.5 deg C per decade which would mean that these areas now are a couple of degrees hotter than 40 -50 years ago.”

    And yet according to GISS they are only 0.5-1.0C hotter than 100 years ago (i.e. 0.05-0.1C/decade). While the other half of Australia is merely 0.2-0.5C hotter than 100 years ago (i.e. 0.02-0.05C/decade)
    http://tinyurl.com/cr98q2

    “To those deniers who may be unaware of the population density of this region, I’d just say that you’d be hard pushed to blame the measured increase on the UHI effect!”

    So let’s say the measured increase for central Australia is ~0.75C warmer than 100 years ago. Some of this is bound to be due to urbanisation around the sites concerned (roads, buildings, other land use changes etc), which can have significant micro-effects even in the smallest settlements. Let’s say, reasonably in my opinion, on the scale of 0.25C. This would leave 0.5C to attribute to AGW in the last 100 years, about the same as what you are claiming per decade.

    Certainly there has been no 0.5C increase for the last decade: compared with 1988-1998, the past decade has been over 0.2C cooler for about a third of Australia towards the NW, over 0.2C warmer for maybe an eighth of Australia in the SE, and flat (within 0.2C) over the remainder.

    http://tinyurl.com/c4wgyj

  • DocNavy // February 12, 2009 at 10:44 pm

    Craig Allen:
    “Where is the global cooling that all the denialists are promising us?”

    Since -as we all know- ANY significant warming event is proof positive of Global Warming, and any significant cooling event is merely anecdotal, short term, background variation, or random in nature; I direct your attention to:
    Anecdotal evidence:

    [edit]

    [Response: I've championed the difference between short-term events and long-term trends, and the statistical significance of same. That's why this post is about trends -- not events. But you're either too lazy to pay attention, too deceitful to care about the truth, or too stupid to understand. Or all three.

    In any case, your list of "anecdotal evidence" is way over the stupid threshold, although it might win you the "cherry-picker of the century" award. If anecdotal evidence is all you care about, have a look at yesterday's record-setting temperatures and note that the hots outnumber the colds by 418 to 39 (yeah, more than ten to one).]

  • TCOis banned...why? // February 12, 2009 at 11:10 pm

    I love Australilia. I went up and down the coast for motnhes as a getting out of the Navy trgyuty and met the sweetest Brit girl ever (they are way friendlier and accessible than the Sheilas).

    Any country that has a decent sized town that is LITERALLY named Surfer;s Paradise. just rocks.

  • Nathan // February 13, 2009 at 12:09 am

    Chris

    “Some of this is bound to be due to urbanisation around the sites concerned (roads, buildings, other land use changes etc), which can have significant micro-effects even in the smallest settlements. ”
    Have you been to Central Australia? There’s no ‘urbanisation’… The aerodromes and stations (that’s Cattle stations) where the data is probably coming from will not have changed… and certainly won;t have asphalt or concrete. It’s all dirt out there.

  • naught101 // February 13, 2009 at 3:41 am

    Tamino: while I understand why you do it, it makes it hard to follow the conversation when you removed slabs of denialist’s posts..

    On the statistical significance of trends, 30 years is often used as a minimum to determine climate trends. Is there a maximum? Or is there an optimum? Or does it all depend on the causes of the changes?

    [Response: There's no set time span; whether a trend is significant or not depends on the "signal-to-noise ratio." If the signal is stronger it takes less time, if the noise is bigger is takes more time. For the present global temperature pattern, 20 years is about the time required, but it's not impossible for noise to obscure the trend even over 20 years, and if the globe starts warming faster then trends will be verifiable in less time than that. For isolated locations, it often happens that the trend is bigger so it takes less time, and it also happens that the noise is bigger so it takes more time.

    As for the "slab" of the denialist's post, it was nothing but a list of cold temperature records for isolated locations. Local high and low records are broken every day; in fact the graph I referred to for yesterday indicated over 400 such record-breaking events in the U.S. alone. More than 90% of those were high-temperature records, but there's so much variation from day to day that sometimes hot records predominate, sometimes cold. As for following the conversation, that requires that there's actually a conversation rather than just blathering.]

  • philscadden // February 13, 2009 at 4:08 am

    I would hate it if my curiousity about the wriggle in the Melbourne was interpreted as “its all caused by ocean oscillations, not global warming”. The comment was sparked by what I had read about Indian Ocean Dipole affecting Australian climate and too busy to actually look up the data. The same report was noting change in IOD (more positives) which might be a response to global warning.

  • Harold Pierce Jr // February 13, 2009 at 7:38 am

    Go over to WUWT and read about the reasons for the brushfires. Briefly, home owners have not been allowed to cut down trees to make firebreaks and the program of controlled burns to remove accumlated fuel of dead plant matter had been greatly curtailed to enviros protests.

    Cutting down 247 trees on his property to create a firebreak ended up costing the home owner $100,000.

    I stongly suspect that many fires wer set by anarchists and other radical enviros.

  • Peter Martin // February 13, 2009 at 10:10 am

    Chris,

    The figures I’m quoting are from the Australian BOM. If I’ve misinterpreted their data maybe you could let me know how.

    Globally, as you might expect due to the thermal inertia of the oceans, there is a greater warming over land areas than over sea. See for example:

    http://www.bom.gov.au/cgi-bin/silo/reg/cli_chg/trendmaps.cgi?variable=tmean&region=aus&season=0112&period=1970

    The change for all land areas being about 1 degC in the last 100 years.

    Most of this has occurred in the last 30 years with warming occurring at ~ 0.3 deg C per decade.

    This rate of warming on land is reasonably consistent with the IPCC ’s estimated range for a doubling of CO2 levels, and in fact could indicate that the true figure somewhat exceeds their mid-point figure of 3 degC.

  • Peter Martin // February 13, 2009 at 10:16 am

    Sorry. Wrong link in previous post.
    Should be: http://data.giss.nasa.gov/gistemp/graphs/Fig.A4.lrg.gif

  • Kevin McKinney // February 13, 2009 at 4:39 pm

    Harold Pierce, your post is both irrelevant to temperature trends and scurrilous.

  • Jim Eager // February 13, 2009 at 10:02 pm

    I see Herald has graduated to baseless slander and libel, not to mention fortune telling.

  • Kipp Alpert // February 13, 2009 at 11:26 pm

    Tamino: That is very intersting how you determined the timeline. I thought it was probably thirty years for understanding climatology, as apposed to WAttsUP five year weather reports. So it’s based on the level of the signal, compared to noise ratio?right?
    KIPP

    [Response: Basically yes. Stronger signal = less time required, stronger noise = more time. While 30 years is a reasonable rough guide, it's not an ironclad rule.]

  • Kipp Alpert // February 14, 2009 at 3:49 am

    cool

  • Johan i Kanada // February 14, 2009 at 6:42 am

    “Stronger signal = less time required, stronger noise = more time”

    Isn’t the much more important aspect the frequency of the signal vs the noise?

    If you find a nice and solid temperature trend during a time period on one year or 5 years, that is hardly proof of anything (climate wise), it doesn’t matter if “the signal” is strong or not. (How do you know what the signal is?)

    A more robust statement would presumably be to say that we are actually looking for long term trends (30 to 50 years) so obviously we need to filter out the higher freequency noise.

    Btw, there is surely an upper limit too, i.e. if you looked for a trend with a T of let’s say 1,000 years (or 10,000 years), then it would not be that relevant for the debate over whether recent CO2 emissions has affected the climate in a significant way or not.

    [Response: To other readers: pay no attention to this Kanada guy. He doesn't know what he's talking about.]

  • Saltator // February 14, 2009 at 12:55 pm

    I have had a closer look at the Melbourne temperature data. My dataset extends from 1948 to December 31 2006.

    Analysis of variance shows that for mean summertime daily maximum temperature, there is no significant difference between decades with the exception of the 2000s vs 1980s (Tukey’s HSD test). It would seem that the increase in temps for Melbourne occur in the other seasons (especially Winter and Spring).

    Bushfire is very closely correlated with temperature. It is also a predominantly summertime phenomenon. Given that, it is unlikely that there has been a significant increase in bushfire potential in Melbourne in the decades from the 1950s.

    A couple of caveats. 1. The dataset is missing 2007 and 2008 data. 2. The increased temperatures for Winter and Spring can lead to increased plant growth and therefore Summer fuel load.

  • Ray Ladbury // February 14, 2009 at 1:03 pm

    Johan, think about it. Imagine you have a strong, persistent signal with short periodicity. You will see pulses where it dominates over time. Now imagine it is constant or very long period–it will always dominate. If the signal drops out for long periods, and your sample is over one of those periods, it would not be a representative sample for the phenomenon.

    Essentially, you aren’t thinking statistically. You can play around with this using Excel. The random number generation routine while kludgy is sufficient to bring out trends.

  • Johan i Kanada // February 14, 2009 at 1:53 pm

    Dear Tamino,
    When it comes to signals & general physics I actually do now what I am talking about.
    But perhaps your not interested in having a “conversation” (as you call it), at least not with me. No problem, it’s your blog.

    [Response: When you post stupid stuff like in your last comment, you yourself prove that you don't know what you're talking about.

    I'd love to have a conversation. But that requires you to get off you high horse and stop positing garbage, just to be contrary. That isn't conversation at all.]

  • Saltator // February 14, 2009 at 1:56 pm

    Very useful site for those interested:

    http://knowledgeweb.afac.com.au/home

  • stewart // February 14, 2009 at 4:41 pm

    Thinking about the Nhill data.
    Nhill is a small country town in Western Victoria, not near any major rivers (unlike Mildura or Deniliquin) and away from the coast (unlike Wilson’s Prom or Cape Otway). I wonder if the declining minimum temperatures are a result of drought - less humidity means more night time cooling. Other sites are more humid, or buffered by large bodies of water. The timing fits with the decrease in rainfall over the past decade as well.

  • Johan i Kanada // February 14, 2009 at 4:44 pm

    Ray,
    “Essentially, you aren’t thinking statistically”
    Correct, I am thinking scientifically ( perhaps with a signal analysis bias).

  • Ian Forrester // February 14, 2009 at 6:09 pm

    JIK, your problem is that you are quite unable to think, whatever way it is construed.

    You spread your ignorant rants all over the internet and claim to have a “science background.” If you really do, you are doing a grave injustice to those of us who really are scientists.

  • Ray Ladbury // February 14, 2009 at 10:44 pm

    Johan, Since when is statistics NOT part of science–or signal analysis for that matter. Have you ever really known a real scientist?

  • Craig Allen // February 15, 2009 at 1:57 am

    Saltator,

    You need to examine your data in more detail.

    This Australian Bureau of Meteorology website lets you noodle through maps of their datasets. You will see that there are clearly differences between the trends in each season, but for most statistics for all seasons in the south-east the trends are bad.
    I live in Victoria, South-eastern Australia. I can assure you that the analyses of the BOM are very much backed up by our anecdotal experience. It is getting hotter and it is getting drier. We watch as year by year our dams and lakes get drier, as our gardens and park-lands parch and die, and as bushfire inducing weather becomes ever more frequent. Last year we had a heatwave that broke records across vast swaths of south-eastern Australia. This year those records were eclipsed. And this is on top of a record drought exceeding a decade in which we are seeing record low inflows to our biggest river system. I come from a farming family on the Eyre Peninsular. The consensus among farmers in my district it that whereas a generation ago there we could expect one good year in three, one OK year in three and one bad year in three, now we get 7 failures in 10 years.

    [Response: You have my sympathy; it must be difficult for everyone in southern Australia, especially these days.

    I would caution that anecdotal evidence can be misleading, and that it's not the events but the trends that define global warming.

    I urge everyone everywhere to remain active in the good times as well as the bad. Wherever you live, even when events are favorable, don't lessen your urging the government and fellow citizens to move swiftly and effectively to renewable energy and carbon neutrality.]

  • JM // February 15, 2009 at 1:53 pm

    Well the “Urban Heat Island” argument is completely out to lunch on all those stations.

    Ballarat, Nhill and Deliniquin are all country towns which have shown no growth over the last 40 years or so (rural populations in Australia are in long term decline), Cape Otway is basically a lighthouse and Wilson’s Promonitory is a large national park that would be lucky to have more than 100 constructions in it total - in fact a small expansion of the only village in the area (Tidal River) a few years ago caused a state-wide political uproar, and was eventually stopped.

    All 5 locations are the exact opposite of urban heat islands and could probably serve as controls.

  • JM // February 15, 2009 at 2:05 pm

    Chris: “Some of this is bound to be due to urbanisation around the sites concerned (roads, buildings, other land use changes etc), which can have significant micro-effects even in the smallest settlements. ”

    Chris let me take you aside for a moment. Australia is the most urbanized country in the world, 85% of our population live in 5 cities on the coast.

    Central Australia is the least populated area of the world after Antarctica - there is no urbanisation there, none. I was once in a town marked on a continent scale map as a major stop on a major highway.

    It was a roadhouse consisting of a sprawling bungaloa, an outhouse, a couple of sheds, a family of five and a talking parrot.

    I was stuck there for 4 days.

    Believe me when I say Central Australia is not “urbanized”, to say it’s “populated” at all is a hell of a stretch.

  • Chris // February 15, 2009 at 5:41 pm

    Yes I have been to Central Aus.

    Remember I was comparing to 100 years ago, when there were no aerodromes, cars etc.

    The label urban heat *island* can be misleading, as can talk of the absolute size of settlements (rather than how they have changed over time). All it takes is for an extra building to go up near the thermometer to make possible the ~0.25C effect I referred to.

    Certainly the temperature dropped away at night when I was there, and it remained warmer close to buildings just like anywhere else in the world.

  • Chris // February 15, 2009 at 6:54 pm

    Here’s a case study re: possible micro-scale “urban” heat effects in central Aus. BOM has weather stations as both Marree and Marree Airport.
    Marree is a town of ~100 population in the north of South Australia. Its weather station is at an altitude of 50m, as is its airport weather station which is 570m away [bonzle.com]

    Marree has readings for 3pm/9am for the last 3 days:

    32.8, 23.8, 31.0, 23.1, 30.1, 22.8 - average 27.3C

    http://www.bom.gov.au/products/IDS60801/IDS60801.94480.shtml

    The equivalent temps for Marree Airport are as follows:

    31.7, 23.6, 30.6, 23.2, 30.2, 22.2 - average 26.9C

    http://www.bom.gov.au/products/IDS60801/IDS60801.95480.shtml

    So micro-scale effects have made Marree on average 0.4C warmer than Marree Airport for that small set of readings.

    Does it prove urban heat effects? Not at all - it’s just a superficial investigation, and could have all kinds of reasons, including random cloud fluctuations.

    Nonetheless, to my mind it indicates that micro-scale urban effects are worthy of consideration.
    I picked Marree because it looks like it may be one of the sites from which the enhanced warming in central Aus has been extrapolated:
    http://data.giss.nasa.gov/cgi-bin/gistemp/gistemp_station.py?id=501944800000&data_set=1&num_neighbors=1

    I don’t want to overstate my point: no doubt siting issues will be something that BOM/GISS have considered. But nonetheless, Marree is classified as a rural area, so presumably unlikely to have had any adjustment.

    [Response: Different locations -- even those a few meters apart -- will show different temperature. That's one of the reasons it's important to work with temperature anomaly rather than temperature. The temperature difference between two places tells us nothing about the trend and nothing about the impact of urbanization; temperatures can be very different even if the trends are identical.

    Your report on the temperature difference between Marree and Marree airport indicates only this: that you're willing to draw a completely unjustified conclusion (about the importance of micro-scale effects) rooted in utter ignorance of the important data (anomalies and trends), based on 6 data points, covering a span of 3 days. You need to improve your critical thinking skills.]

  • Chris // February 15, 2009 at 8:28 pm

    “you’re willing to draw a completely unjustified conclusion”

    I know something of micro-scale effects from my studies including a recent meteorology project, and papers such as Hinkel et al confirm that UH effects can be surprisingly large for small settlements (albeit that was a study of a high-latitude site in winter). I wouldn’t say I’m drawing a conclusion (sorry if I’m implying that), more giving an illustration, the point being that if differences in micro-scale effects can result in two locations 500m apart having a consistent differential in temperature, then *changes* in micro-scale effects for a given location can also result in a differential in temperature between the location as it is and as it would have been without the changes.

    [Response: You must think very little of the intellect of the readers here. The whole point of your comment was to suggest that urbanization had caused a difference in such a small place as Marree. Despite including caveats, you also drew a conclusion: "So micro-scale effects have made Marree on average 0.4C warmer than Marree Airport for that small set of readings." The implication is clear.

    What's even more clear is that anybody who draws a conclusion, or even makes a suggestion of same, based on 6 readings covering 3 days for 1 location, had better not call himself a "skeptic." You're as gullible as it gets. To defend yourself with "I wouldn't say I'm drawing a conclusion (sorry if I'm implying that), more giving an illustration" is disingenuous.]

  • Chris // February 15, 2009 at 8:44 pm

    “You must think very little of the intellect of the readers here”

    Well I’ve made my points in good faith. You’ve made your opinions about them clear, no doubt each reader will have his/her own opinions, if they agree with you then so be it.

  • luminous beauty // February 15, 2009 at 9:37 pm

    Chris,

    Mathematical significance is not a matter of opinion.

    Good faith + bad math = nonsense

  • Hank Roberts // February 15, 2009 at 10:10 pm

    New Scientist blogging from the AAAS meeting:

    Climate models predicted Australian bushfires
    * 15 February 2009 * Magazine issue 2695.

    AUSTRALIA may have just had a horrifying preview of what climate change has in store for its people. Even early warning couldn’t stop last weekend’s bush fires in Victoria claiming 170 lives and over 700 homes.

    Climate models based on figures from the Intergovernmental Panel on Climate Change predict more frequent - and more extreme - fires for southern Australia over the next few decades. Yet the role of climate change in recent fires has been downplayed, suggests John Handmer of the Bushfire Cooperative Research Centre at RMIT University in Melbourne….

  • Chris O'Neill // February 16, 2009 at 2:49 am

    Saltator:

    Analysis of variance shows that for mean summertime daily maximum temperature, there is no significant difference between decades with the exception of the 2000s vs 1980s

    According to the Australian Bureau’s trend map, the 59 year trend in Melbourne was about 0.1 deg C/decade. I take everything Saltator says with a grain of salt after he falsely claimed that Melbourne’s recent record highs had something to do with an urban heat island so Tamino, could you work out for us if Melbourne’s temperature trend of the last decades is “statistically significant”.

    Bushfire is very closely correlated with temperature. It is also a predominantly summertime phenomenon. Given that, it is unlikely that there has been a significant increase in bushfire potential in Melbourne in the decades from the 1950s.

    What? Even if the temperature rise hasn’t been “significant” (whatever that means), haven’t you heard of rainfall? Victoria is much drier since 1997. Isn’t bushfire risk correlated with rainfall?

  • Richard Steckis // February 16, 2009 at 3:17 am

    Hank. Climate models did nothing of the sort. You tell me which model predicted the Melbourne bushfires of 2009.

    You people hang too much on climate modelling and climate model prediction. Tell me which run of which model with which parameterizations at which institution with which climate data and how it was used, that produced a miraculous prediction that Melbourne Australia would experience the third or fourth most serious fire on record in 2009.

    The answer is none.

  • Richard Steckis // February 16, 2009 at 3:51 am

    “Even early warning couldn’t stop last weekend’s bush fires in Victoria claiming 170 lives and over 700 homes.”

    What early warning?

  • Saltator // February 16, 2009 at 3:52 am

    Richard Steckis should read Saltator. Need to use nom de plume for certain reasons.

  • Saltator // February 16, 2009 at 4:04 am

    Chris O’Neil,

    I analysed mean summertime daily maximum temperature. That is the critical time and temperature to consider for bushfire occurrence. Rainfall is not highly correlated with wildfire. In fact less rainfall should mean less growth and therefore a lower fuel load the following summer.

    I am saying there has been no significant increase in the critical summertime maximum temperatures.

    Interestingly, the worst decade for fires in Victoria was the 1980s which just so happen to have the coolest summer maximums of any decade from the 1950s.

  • Hank Roberts // February 16, 2009 at 4:06 am

    Sorry, RS, I quoted a brief excerpt you can find in New Scientist. You’ve made up a claim that’s not in the article, and then denied it could be possible.

    Citing to your own imagination? Why not do the work of looking up the article …..

    Oh, why bother.

  • Ian Forrester // February 16, 2009 at 4:19 am

    Steckis, you have gone completely off the rails. Climate models predict climate not things that are caused by climate.

    By the way, why has my recent comment on your blog remained in moderation for about four days now?

    [Response: I haven't seen any comment in moderation that long, or even close. There must be a bug in the system. Or are you referring to saltator's blog?]

  • Hank Roberts // February 16, 2009 at 4:19 am

    Oh heck, there’s always going to be some youngster needing help with a term paper. A few places to start. As always, read the citing articles and the references/footnotes.

    http://www.springerlink.com/content/r5402518r58mx051/

    http://books.google.com/books?hl=en&lr=&id=4f8anIoS33MC&oi=fnd&pg=PA26&dq=%22Climate+model%22++%22Intergovernmental+Panel%22+%22Climate+Change%22++%2Bfires+%2BAustralia&ots=kDeoN_u2jj&sig=bcUz6GEnnvibqTc_UbCcfhPGuu4

    Those aren’t definitive answers; just examples.

    Anyone at a university has a librarian available to help look up the original New Scientist article and follow up the information therein.

    Enjoy.

  • chriscolose // February 16, 2009 at 4:19 am

    Richard Steckis,

    For one who seems so confident in his understanding to insult Hank and others, you are quite ignorant about the subject of climatology and what models are actually being used for. It is fine to make useful criticisms about the ability of current generation AOGCM’s to make forcasts in the future, but if you can’t understand the fundamental concept of what people are actually trying to predict, then you’re not in a position to do so.

  • JM // February 16, 2009 at 4:46 am

    Saltator: “In fact less rainfall should mean less growth and therefore a lower fuel load the following summer.”

    It might help to understand the nature of Australian vegetation before going off on psuedo plausible flights of fancy.

    1. Euclypts do just fine in dry climates so lack of rain doesn’t slow their growth much. They also drop leaves continuously and also branches and twigs - the fuel load on the ground is replenished throughout the year. This is why we do burn offs in Australia and land owners go around their houses removing all the dead vegetation.

    2. Rainfall has substantial impact on how flammable all that dead stuff is. When I was young I used to visit the Dandenong ranges frequently and the ground fuel was often sodden wet. After 12-15 years of drought that is no longer true, it is tinder dry.

  • dhogaza // February 16, 2009 at 5:23 am

    Or are you referring to saltator’s blog?

    Duh :)

    Sorry, I think it was a clear statement …

  • Ian Forrester // February 16, 2009 at 5:35 am

    Tamino, I was referring to a post I made on Steckis’ blog.

    Sorry for the confusion.

  • Saltator // February 16, 2009 at 6:36 am

    Hank. Your thread stated:

    “New Scientist blogging from the AAAS meeting:

    Climate models predicted Australian bushfires
    * 15 February 2009 * Magazine issue 2695.

    AUSTRALIA may have just had a horrifying preview of what climate change has in store for its people. Even early warning couldn’t stop last weekend’s bush fires in Victoria claiming 170 lives and over 700 homes.”

    Does that not say that the models predicted the fires?

    JM,

    Humidity is more important for bushfires. I grew up in rural Western Australia and am cognisant of eucalypt growth dynamics.

    Of course southern Australia has dry hot summers. Bushfires are largely confined to summer and early autumn. Rainfall is lowest at these times. The critical factor for fires is the moisture content of the fuel load (ground leaf litter or dry grass which is the primary source of fuel). The moisture content of those fuel sources is controlled by humidity. Therefore, for bushfires, the most important factor is dewpoint temperature and not rainfall.

  • Saltator // February 16, 2009 at 6:37 am

    Oh. And Hank. What early warning? South east Australia does not yet have an early warning system in place for bushfire.

  • Saltator // February 16, 2009 at 6:43 am

    And JM,

    Eucalypts will grow when they have available water and some warmth. That is different than saying they can grow fine in dry climates. That is only true if there is water available to enable growth. That is probably why in the extreme desert situations you will most likely find them along watercourses where available groundwater allows them to survive.

  • Chris O'Neill // February 16, 2009 at 9:24 am

    Richard Steckis:

    Tell me which run of which model with which parameterizations at which institution with which climate data and how it was used, that produced a miraculous prediction that Melbourne Australia would experience the third or fourth most serious fire on record in 2009.

    Just another person who doesn’t want to understand the difference between weather and climate.

  • Saltator // February 16, 2009 at 9:49 am

    Ian Forrester,

    The statement was made by Hank that the models predicted the fires in Australia. I am fully aware of what the climate models are or are not supposed to predict.

  • Chris O'Neill // February 16, 2009 at 11:28 am

    Saltator:

    I analysed mean summertime daily maximum temperature.

    So did the Australian Bureau of Meteorology.

    That is the critical time and temperature to consider for bushfire occurrence. Rainfall is not highly correlated with wildfire.

    Citation please.

    In fact less rainfall should mean less growth and therefore a lower fuel load the following summer.

    Growth in Victoria is temperature-limited in and around winter, so summer rainfall is not the only thing that determines fuel load.

    I am saying there has been no significant increase in the critical summertime maximum temperatures.

    Anyone can make an assertion. We already know the summertime maximum temperatures have trended upwards. I have yet to see a demonstration that this trend is insignificant.

    Interestingly, the worst decade for fires in Victoria was the 1980s

    Citation please.

    which just so happen to have the coolest summer maximums of any decade from the 1950s.

    The Australian Bureau of Meteorology disagrees with you. The coolest average of summer maximums was the 1950s while the 1980s had the second warmest average of summer maximums with the 2000s (9 summers) being the warmest by far.

    I’m beginning to get rather tired of your false assertions. You may not care but it makes you look consistently biassed and not very credible.

  • Ray Ladbury // February 16, 2009 at 11:43 am

    Richard/Saltator, you seem to have adopted the A. Watts school of analysis–namely, remain maximally ignorant of the phenomenon you are analyzing so as not to spoil your objectivity. Then use your intuition to imagine what the phenomenon must be like.

    News bulletin: It ain’t workin’.

  • Kevin McKinney // February 16, 2009 at 2:30 pm

    Here’s the relevant excerpt from the New Scientist story that Hank referred to:

    “Yet the role of climate change in recent fires has been downplayed, suggests John Handmer of the Bushfire Cooperative Research Centre at RMIT University in Melbourne.

    Certainly, last weekend’s fires were unprecedented: “We had a record heatwave, the worse fire danger index on record, during a record-breaking drought,” says Handmer. The fire danger index takes into account both temperature and humidity. Over 50 is extreme; on Saturday, the index is believed to have been 5 to 6 times higher.” [Handmer is identified as a staff scientist with the Bushfire Research Cooperative.]

    I think we can stop talking about whether or not the weather conditions were unusually conducive to bushfires.

  • saltator // February 16, 2009 at 4:55 pm

    Kevin:

    “I think we can stop talking about whether or not the weather conditions were unusually conducive to bushfires.”

    Of course they were. I am not disputing that the fires of January 2009 occurred in exceptional weather circumstances. That doesn’t mean that it is evidence of a global warming event.

  • saltator // February 16, 2009 at 5:11 pm

    Chris,

    The graph you cite clearly shows the 1980s as a decade having substantial negative summer maximum anomalies. My analysis was for Melbourne Regional Office. Your data are for Victoria as a whole. I also did state that the 2000s were higher than the other decades but that the 2000s were only significantly higher than the 1980s. I would be happy to provide my R script if you wish to peruse it yourself.

    I make the statement about the 1980s having more bushfires over the decade from the report to the Premier and Cabinet of the 2002-2003 bushfires Appendix IV Figure IV.2 This appendix can be obtained here: http://www.oesc.vic.gov.au/wps/wcm/connect/OESC/resources/file/eb071c0e4a3828c/Appendix%20_IV.pdf

  • saltator // February 16, 2009 at 5:26 pm

    Chris:

    “Anyone can make an assertion. We already know the summertime maximum temperatures have trended upwards. I have yet to see a demonstration that this trend is insignificant”

    It was not an assertion Chris. It was the result of an analysis of variance of mean summertime maximum temperature by decade. The Tukey’s HSD test showed that the only significant difference was that between the 1980s and 2000s no other comparison was significant.

    If that an ANOVA is not a proper demonstration then what is? Tamino has had ample time to check my results. He has not dismissed them yet so I can only assume they are robust.

    [Response: I didn't check your results. But the fact that decades do not show a significantly different average, is not the same as that there is no trend.

    How did you define "summer"? Is that Dec-Jan-Feb?]

  • saltator // February 16, 2009 at 5:56 pm

    Tamino.

    True, I agree that averages do not show a trend but I was not trying to do trend analysis but to see if there was a significant difference in temperature between decades. Perhaps I did not answer Chris’ statement clearly enough that I was not looking at trends.

    I defined summer as Dec to February.

  • Kevin McKinney // February 16, 2009 at 7:48 pm

    Saltator,

    What we know:

    1) There is a significant warming trend for Melbourne & environs, as demonstrated in the original posts;
    2) The fires resulted from weather conditions considered highly unusual in historical terms;
    3) The fires occurred in the context of a 12-year drought;
    4) Climate modeling has long predicted increased drought and more extreme weather events as a consequence of AGW. (I don’t know if there were predictions more specific to the region, as the New Scientist story seemed to suggest. I suspect that bushfire risks were (and are) modeled by dedicated software driven by data derived from GCMs, but don’t have specific knowledge that this is what was meant in the report.)

    Whether there is something here we can strictly term “evidence of a global warming event”–I assume you mean in specific connection with the Melbourne tragedies, as there is already ample evidence of global warming in general–I leave to our good host.

    But the overall picture certainly does nothing to suggest that climate complacency is a wise choice.

  • JM // February 16, 2009 at 9:11 pm

    “Eucalypts will grow when they have available water and some warmth. That is different than saying they can grow fine in dry climates. That is only true if there is water available to enable growth. That is probably why in the extreme desert situations you will most likely find them along watercourses where available groundwater allows them to survive.”

    Apart from being blindingly obvious, I’m not sure what you are arguing.*

    You appear to be arguing that trees dissappear in droughts yet we’ve just had bushfires where still-existent trees burnt.

    Those not-non-existent trees were set alight by (initially) underbrush fires promoted by the presence of a fuel load that was tinder dry due to lack of rain over a period of more than a decade.

    You have argued explicitly that the fuel load shouldn’t (in your opinion) be there due to low rainfall.

    What’s your point? That the recent fires are a figment of our imagination because the drought made them impossible?

    * Trees get their water from the ground, euclypts often have deep roots and there is a lot of ground water in Victoria. My point about rainfall was the flammability of ground material during droughts.

  • David B. Benson // February 16, 2009 at 9:20 pm

    There is now a guest thread about the southeast Australia fires on RealClimate:

    http://www.realclimate.org/index.php/archives/2009/02/bushfires-and-climate/langswitch_lang/fr

  • Saltator // February 17, 2009 at 5:59 am

    http://www.ga.gov.au/hazards/bushfire/causes.jsp

    and JM don’t be like Dhogaza and put words in my mouth. I did not say the fuel load would not be there. I did not say that the trees will disappear in a drought.

    Those are your words not mine.

    You are right. Rainfall is an important factor.

  • Chris O'Neill // February 17, 2009 at 10:38 am

    Saltator:

    The graph you cite clearly shows the 1980s as a decade having substantial negative summer maximum anomalies.

    What you said was:

    Interestingly, the worst decade for fires in Victoria was the 1980s which just so happen to have the coolest summer maximums of any decade from the 1950s.

    Pity you forgot to mention which averages you were talking about. You obviously think differently but I don’t find it particularly interesting that a decade with a couple of cooler summers had big bushfires during the other summers.

    My analysis was for Melbourne Regional Office. Your data are for Victoria as a whole.

    Well you were talking about bushfires in *Victoria*.

    I also did state that the 2000s were higher than the other decades but that the 2000s were only significantly higher than the 1980s.

    For Victoria as a whole (which is probably more relevant to Victorian bushfires than is Melbourne city), the 1980s were at least slightly warmer than all the other decades except for the 2000s. If the 2000s are significantly warmer than the 1980s, I don’t know how to avoid concluding that the 2000s are significantly warmer than the 50s, the 60s, the 70s, the 80s and the 90s.

    I make the statement about the 1980s having more bushfires over the decade from the report to the Premier and Cabinet of the 2002-2003 bushfires

    OK, so that report is out-of-date as far as the 2000s go. Since then the 2000s have had the massive 2006 fires and of course the fires this year. The 2000s look like being the biggest decade for fires by a long way. Correlates well with the 2000s being the warmest decade by a substantial amount (at least 0.6 deg C warmer than any other decade since 1950).

  • saltator // February 17, 2009 at 12:44 pm

    “You appear to be arguing that trees dissappear in droughts yet we’ve just had bushfires where still-existent trees burnt.”

    did I? Where?

    “You have argued explicitly that the fuel load shouldn’t (in your opinion) be there due to low rainfall.”

    I did not argue any such thing either explicitly or implicitly. I argued about fuel load being reduced. A very different thing.

    However, you are right. Rainfall is important.

  • saltator // February 17, 2009 at 5:04 pm

    “If the 2000s are significantly warmer than the 1980s, I don’t know how to avoid concluding that the 2000s are significantly warmer than the 50s, the 60s, the 70s, the 80s and the 90s.”

    The analysis of variance does not support you contention. The aposteriori test shows that the only significant difference was between the 2000s and the 1980s. No other test between decades was significant.

    I have also run the analysis with Ballarat data. The results are similar except the 2000s and 1970s are significantly different with no other test being significantly different.

    Do you want the R script to confirm?

    [Response: What you've done is limit your analysis to such small spans of data that there's barely any chance for a statistically significant result. You also haven't said anything about the uncertainty in the results. If the difference between the 1990s and the 2000s is 0.7 plus or minus 1.2 deg.C, does that mean there's no change so we shouldn't worry at all? Or does it mean that the change in one decade could be as high at 1.9 deg.C so we should panic? Or -- maybe -- does it mean that your analysis is based on such limited sample sizes that it gives us no useful information at all?

    With such small subsamples, the fact that any of the results at all turned out significant is rather startling testimony to the magnitude of the change.]

  • saltator // February 17, 2009 at 5:17 pm

    Chris. Please be mindful that I looked at daily summertime max temperatures. Reason being, summer is bushfire season.

  • JCH // February 17, 2009 at 5:22 pm

    I did read that spring 2008 rains encouraged new undergrowth, which had since died and dried out.

  • Saltator // February 18, 2009 at 1:20 am

    “Response: What you’ve done is limit your analysis to such small spans of data that there’s barely any chance for a statistically significant result. You also haven’t said anything about the uncertainty in the results. If the difference between the 1990s and the 2000s is 0.7 plus or minus 1.2 deg.C, does that mean there’s no change so we shouldn’t worry at all? ”

    The data are 90 values per year making up 900 values per decade. Are you saying that this is too small a sample size?

    [Response: I used the "summer" data from the BOM website, which gives only 10 values per decade but still comes close to significance. It appears you're using daily data, in which case: did you use anomalies rather than temperatures? With daily data, failing to take anomalies first will dramatically inflate the variance.]

  • Saltator // February 18, 2009 at 2:00 am

    I converted the temperature to anomalies by deducting the average GISS baseline temperature of 14C from each value. Could that be a problem? Am I being too simplistic using that value?

    [Response: That doesn't convert them to anomaly, it just shifts the zero point. The purpose of computing anomaly is to remove the seasonal pattern, so you have to subtract from each data value the average for that same date, for that same location.]

  • Saltator // February 18, 2009 at 2:08 am

    Also. I am using data published in the SEACI website:http://www.mdbc.gov.au/subs/seaci/research_data.html

    Regards

  • Saltator // February 18, 2009 at 5:45 am

    “Response: That doesn’t convert them to anomaly, it just shifts the zero point. The purpose of computing anomaly is to remove the seasonal pattern, so you have to subtract from each data value the average for that same date, for that same location.”

    Ok. Back to the drawing board. Thanks.

  • dhogaza // February 18, 2009 at 7:05 am

    Ok. Back to the drawing board. Thanks.

    Nice to see some progress here. Wasn’t that long ago that you’d claim to have overturned all of climate science.

    Now it appears that you believe that you might have to learn some things before you can overturn all of climate science.

    That’s real improvement.

    Perhaps someday you’ll learn that there are axiomatic understandings in climate science that are based on well-proven physics and that no matter how hard you try, you’re not going to overturn this.

    Meanwhile … please continue to try. Don Quixote is a famous Spanish novel for good reason.

  • Chris O'Neill // February 18, 2009 at 2:39 pm

    Saltator:

    Please be mindful that I looked at daily summertime max temperatures. Reason being, summer is bushfire season.

    That’s what I thought you were looking at.

  • Saltator // February 19, 2009 at 1:04 am

    “Perhaps someday you’ll learn that there are axiomatic understandings in climate science that are based on well-proven physics and that no matter how hard you try, you’re not going to overturn this.”

    Axiomatic systems may be fine for mathematics. However they do not do particularly well in a “real world” situation where the system is subjected to many different influences. Please remember, the well proven physics you refer to is not the exclusive privy of the pro-AGW lobby.

    [Response: Get real. Anti-AGW advocates consistently ignore, abuse, and violate well-proven physics. Monckton.]

  • Saltator // February 19, 2009 at 2:14 am

    Unless by axiomatic you meant self-evident. I do not believe that current research makes anything “self-evident”.

  • Hank Roberts // February 19, 2009 at 3:53 am

    Have you ever used a laser?

  • dhogaza // February 19, 2009 at 5:41 am

    Have you ever used a laser?

    Sssshhhhh CO2 lasers are a commie lie that don’t really exist but ssssshhhhh don’t tell no one, OK?

  • JM // February 19, 2009 at 6:25 am

    Saltator, I think he means axiomatic in the same sense that everyone else does, namely “these are the fundamental assumptions of my model that I’m not going to bother to prove because

    a.) they are well understood and proven by other people in other contexts and

    b.) they involve low level details that don’t have any effect on my model.

    Euclid’s axioms cannot be proven within Euclid’s geometry so they are assumed. They are external properties without which no system can be built or understood.

    Axiom does not mean self-evident, it just means assumed.

    There is nothing self evident about parallel lines for example, Reimannian geometries violate that assumption. But so long as the assumption is valid Euclid’s model (system) works.

    Similarly, a carpenter has an understanding of wood in terms of grain and relative hardness - he doesn’t bother to “prove” his assumptions in terms of cell structure or biology.

    In the same way, discussions of climate science assume the results of the last 150 years of radiative (cf. lasers, computers, atomic bombs etc) and thermal physics because the principles are very well proven by others and it is pointless to reestablish them.

    Or to try to disprove or question them either, which is what dhagoza is accusing you of attempting to do. As it happens, I agree with him.

    You, along with quite a few other denialists, do appear to be attempting to do just that on occasion.

    Bottom line:-

    (1) the radiative properties of CO2 are very well proven and yes it does absorb and emit infra-red.

    (2) the character of re-emmission and it’s thermal consequences (the greenhouse effect) are also very well known and proven.

    You have absolutely no hope of overturning those conclusions and you make yourself look foolish when you try.

  • Saltator // February 19, 2009 at 2:06 pm

    Response: Get real. Anti-AGW advocates consistently ignore, abuse, and violate well-proven physics. Monckton.

    Rubbish. I don’t know about Monckton but most scientists even Spencer and Christie stay within the bounds of the laws of physics. They would not get published in AGU publications and other front line journals if they didn’t. Spencer himself has used GCMs (which utilise the laws of physics) in his published research. I think you need to get over your hatred of these people and look at their research as a true skeptic would. With an open but skeptical mind.

    [Response: The point is this: of course the laws of physics aren't the sole province of AGW believers, but utter disregard for them is solely the province of denialists. As for Spencer, his attachment to creationism, and his idiocy about CO2 increase not being entirely anthropogenic, are abundant testimony to his willingness, even eagerness, to be in denial not just about climate, but about much more fundamental scientific issues. The treatment of well-established sound science by denialists is creepy.]

  • Saltator // February 19, 2009 at 2:07 pm

    That’s all I am going to say on that one as it is OT.

  • Ray Ladbury // February 19, 2009 at 3:22 pm

    Saltator, In this case, axiomatic means that they are a starting point for any serious treatment of the subject. You simply cannot understand the subject without them. The role of CO2 in climate fits the bill. The physical properties are right, it is both long-lived and well mixed, and as such, it’s signature in paleoclimate and the current climate sticks out like a sore thumb. The axiomatic nature of the role of CO2 is evident in the models–not one of which can accommodate a sensitivity less than 2 degrees per doubling.

    As to Spencer and the other denialists. Well, I am reminded of the famous Andrew Lang quote: “He uses statistics as a drunkard uses a lightpost–for support, rather than illumination.” It’s one of the reasons why Spencer hasn’t contributed anything of value to our understanding of Earth’s climate. You see the same thing in the denialosphere over these fires in Oz. They will do anything to divert attention away from the likely role climate change plays in increasing susceptibility to such disasters.

    See, using a Marlin Perkins-esque transition, you can bring the discussion back to the topic.

  • dhogaza // February 19, 2009 at 4:28 pm

    Spencer himself has used GCMs (which utilise the laws of physics) in his published research

    But he ignores the fossil fuel isotope fingerprint that makes it absolutely uncontroversial that burning fossil fuels is responsible for observed increases in CO2 concentrations.

    That’s about as whacked out as you can get in climate science.

  • Gator // February 19, 2009 at 7:46 pm

    Saltator, you should look very closely at the huge gap between what Spencer and Christy get published in reputable journals and what they say in public. In the journals, they stick to what is supported by data. In public, they spout easily refuted nonsense. From people who should know better this might be termed “lying.”

  • t_p_hamilton // February 19, 2009 at 9:15 pm

    I went to a talk by Christy, and one of the things he brought up “against” AGW in response to a question by a little old lady was the 800 year time lag between warming start and CO2 increase in paleoclimate, posing it as a question. It was really pathetic when that elderly amateur knew the answer right away, and Christy didn’t.

  • David B. Benson // February 19, 2009 at 10:07 pm

    t_p_hamilton // February 19, 2009 at 9:15 pm — How did her answer go?

  • dhogaza // February 19, 2009 at 10:15 pm

    It was really pathetic when that elderly amateur knew the answer right away, and Christy didn’t.

    Christy knows the answer, I’m sure, which makes it sadder

  • Ray Ladbury // February 20, 2009 at 12:40 am

    t_p_hamilton says, “I went to a talk by Christy…”

    Why?

  • t_p_hamilton // February 20, 2009 at 4:51 am

    The little old lady said the warming was started by orbital cycles. The reason I went is because our University has a discussion book, this year it was Field Notes from a Catastrophe. There are various seminars through the year related to the book .

  • Former Skeptic // February 20, 2009 at 5:31 am

    OT, but I went for Christy’s talk last month at the AMS annual meeting - yes, the same one where Jim Hansen received the Rossby award that led to Bill Grey going all berserk:

    http://ams.confex.com/ams/89annual/techprogram/paper_143637.htm

    IMO he tried the “woe is me approach vs. the big bad Hansen,” but it wasn’t very convincing to the audience…

  • RedLogix // February 22, 2009 at 7:43 pm

    Interestingly the Australian authorities are now realising that in addition to the 201 tragic deaths directly the result of wildfires, they may well have had a similar number of excess deaths simply due to the heatwave itself.

    http://www.stuff.co.nz/4856060a12.html

    This is very reminiscent of the huge death toll in France a few years ago, and is a reminder that it is the extreme events that count in terms of survival.

    Some stations in Victoria were hitting close to 48 degC… that is hellish hot. It’s not too far off unthinkable temperatures in the 50’s, at which point the possibility widespread electrical system failures, loss of airconditioning and mass deaths of whole populations would no longer be irrsesponsible scaremongering.

    Anyone got any idea just what the upper boundaries of human temperature survivability are? I’ve personally worked in some very high temperatures in industrial settings, and even as a young man found it very debilitating… what would happen for instance if a whole city had a heatwave for 2-3 days at a sustained 50degC, and the power went off?

    (PS most industrial control system components are rated to 50degC ambient, but due to heat production within their enclosures, their actual ambient would usually be 5-10degC hotter than outside temperature. In other words these very hot weather events are stressiing vital infrastructure components to the max and beyond.)

  • Ray Ladbury // February 23, 2009 at 1:23 am

    Redlogix, Can’t speak for a lot of industrial equipment, but most commercial electronics is rated up to ~75 degrees C, 85 at the outside. The military temperature range is up to 125 degrees C. However, reliability falls dramatically once you get much above 50 degrees C.

  • Barton Paul Levenson // February 23, 2009 at 10:25 am

    Dole (1964) considered a planet habitable if 10% of its surface or more had mean annual temperatures in the 273-303 K range (0-30 C), with seasonal averages 263-313 (-10-40 C). 48 C is definitely pushing it.

    Ref: Dole, Stephen H. 1964. Habitable Planets for Man. NY: Blaisdell Press.

  • Saltator // February 24, 2009 at 5:01 am

    BPL,

    48C is not an average. You can have extremes in an average. It is unlikely that Victoria will see many such days in the coming century.

    RedLogix,

    There are nearly double the number of deaths from extreme cold than from extreme heat. see: http://www.csccc.info/reports/report_23.pdf

    read table 3. and Figure 5.

    Death rates due to extreme weather events have substantially declined in the last two decades.

    [Response: Are you really such a sucker as to fall for that "more deaths from cold than heat" crap? The danger from global warming is lack of FOOD and WATER. In the past, civilizations have collapsed from climate change, but not because they couldn't keep warm at night or cool during the day.

    You fell hook line and sinker for on of the dumbest arguments on record.]

  • JCH // February 24, 2009 at 1:49 pm

    “There are nearly double the number of deaths from extreme cold than from extreme heat. …”

    Wow, if we could just have year-round summer, the average life span might reach the moon. The benefits of CO2 pollution are just astonishing.

  • Kevin McKinney // February 24, 2009 at 5:53 pm

    Saltator, this idea regarding the comparative lethality of heat vs. cold events is often stated, as you have done, as fact. I appreciate that you provided a citation to back up your assertion. However, I note that:

    1) The report is from an advocacy group using the term “alarmist” with reference to climate science, hence begin their study with a partisan agenda, and

    2) The conclusion you reference is, to say the least, highly dependent on their chosen methodology. Specifically, they counted heat-related deaths as such *only* if heat was specifically noted on death certificates. Seems like a good recipe for an undercount to me.

    Here’s what the NCDC has to say on the topic:

    http://www.weather.gov/om/hazstats.shtml

    Your study group decided this was less authoritative/reliable than the CDC estimates. I feel otherwise, based on my personal experience: blankets are much cheaper and more reliable than AC.

    At the least, however, there seems good reason to view this as an unsettled question.

  • Saltator // February 25, 2009 at 1:39 am

    [Response: Are you really such a sucker as to fall for that "more deaths from cold than heat" crap? The danger from global warming is lack of FOOD and WATER. In the past, civilizations have collapsed from climate change, but not because they couldn't keep warm at night or cool during the day.

    You fell hook line and sinker for on of the dumbest arguments on record.]

    Sometimes you are too quick on rhetoric and short on facts. There is no current or projected food production crisis (at least to 2030).

    see: http://www.fao.org/docrep/004/y3557e/y3557e00.htm

    http://www.fao.org/english/newsroom/news/2002/7828-en.html

    Your comments are reminiscent of Paul Erlich’s failed prophesies of world food shortages:

    “The battle to feed humanity is over. In the 1970s the world will undergo famines . . . hundreds of millions of people (including Americans) are going to starve to death.” (1968)

    Yes. Past civilisations have collapsed from climate, actually more precisely weather, changes (Mesopotamia, The Moche etc). However, the major proportion of the people did not perish, just the organised societies in which they lived. The people moved or re-organised their societies (research the Moche).

    The problem is not food production but food security (http://en.wikipedia.org/wiki/Food_security#Global_water_crisis)

    [Response: Yes, you really are such a sucker that you fell for that "more deaths from extreme cold than extreme heat" crap.]

  • dhogaza // February 25, 2009 at 4:58 am

    However, the major proportion of the people did not perish, just the organised societies in which they lived.

    Oh, gosh, this is a good reason to continue to increase our fossil fuel consumption.

    All we have to worry about are that our organised societies will perish.

    Thanks for the calming info, there.

  • Saltator // February 25, 2009 at 8:34 am

    [Response: Yes, you really are such a sucker that you fell for that "more deaths from extreme cold than extreme heat" crap.]

    When the truth hurts you resort to abuse I guess. Produce the contradictory evidence (i.e. cite please).

    And where is my subsequent post on increasing aridity being the impetus for civilisation?

    [Response: If you don't want to be treated like an ass, stop behaving like one. You produced that "more deaths from cold than heat" crap for the same reason you went on and on about that stupendously idiotic "non-anthropogenic component to CO2 increase" crap, because you want to discredit the need to act on global warming. Now you want to change the subject and act like you're looking for some kind of productive dialogue. You're just trying to cover your embarrassed ass.

    If you really want productive dialogue, Step 1: stop the propaganda campaign. Step 1, part 1: admit it to yourself.

    Maybe then we can talk. Until then, frankly, I'm sick of you.]

  • Chris O'Neill // February 25, 2009 at 10:46 am

    Saltator :

    48C is not an average. You can have extremes in an average. It is unlikely that Victoria will see many such days in the coming century.

    I should hope not. There weren’t any AT ALL last century.

  • Barton Paul Levenson // February 25, 2009 at 12:13 pm

    Saltator,

    I know the 48 C measurement wasn’t an average. I’m not saying it was, but I can see how you could have gotten that from my post, which was poorly phrased. I would state my argument as: the global annual extreme for habitability is an average of 303 K, the seasonal extreme is 313 K, and 48 C is 321 K, so pushing it if it were to continue even for a few weeks. I know it was just one day. Sorry I wasn’t clearer.

  • Kevin McKinney // February 25, 2009 at 1:40 pm

    “Paul Ehrlich was wrong in 1970, therefore you are wrong today.” Logic 101, anybody?

    As to current food production & climate, how about this:

    http://www.guardian.co.uk/society/2008/aug/11/aid.africa

    (I will give Saltator some credit, however, for adroitness in changing the subject. The name does mean “jumper,” right?)

  • Kevin McKinney // February 25, 2009 at 1:47 pm

    I will add that the Wiki article on food security (really, I ought to say, food INsecurity) is pretty scary–not the optimum start to anyone’s morning. It’s hard to believe he read it, especially the sections relating to climate change and food security.

    Following up the bibliography a bit, the 2009 study (actually, it turns out, a whole new interdisciplinary journal) mentioned came out at the beginning of the month and is available in its entirety for free. The link is here:

    http://www.springerlink.com/content/1876-4517?sortorder=asc&v=condensed

    The last article, on the African Food Revolution, is actually encouraging, though perhaps not to doctrinaire laissez-faire market types.

  • JCH // February 25, 2009 at 4:08 pm

    “more deaths from extreme cold than extreme heat”

    Yes, just finished reading The Diseases of Antarctica. Lord help you mates if those nasty buggers were to start moving north.

    It’s hilarious how fooled some people can get themselves.

  • saltator // February 25, 2009 at 8:19 pm

    “If you really want productive dialogue, Step 1: stop the propaganda campaign. Step 1, part 1: admit it to yourself.”

    It was merely a response to another post about expected deaths from extreme heat. It was meant to bring a factual balance. Nothing more.

    I am not on any propoganda campaign on GW. I know full well that GW is occurring. It is the A bit that I am skeptical about. I will call into question so-called self evident beliefs (that is what a skeptic does). Sorry if that offends.

  • Hank Roberts // February 25, 2009 at 8:51 pm

    The heat/cold thing is much studied. Don’t rely on PR blogs for good scientific information.
    http://scholar.google.com/scholar?sourceid=Mozilla-search&q=death+toll+heat+wave+calculations

  • David B. Benson // February 26, 2009 at 12:34 am

    saltator // February 25, 2009 at 8:19 pm — The A bit has been going on for a long time. Read paleoclimatologist W.F. Ruddiman’s “Plows, Plagues and Petroleum” and then his more recent papers, available from his web site.

  • dhogaza // February 26, 2009 at 12:51 am

    I will call into question so-called self evident beliefs (that is what a skeptic does).

    What makes you think “A” is held as a “self-evident belief”? You think an entire world of professional working scientists are publishing papers, generating working models, refining our understanding of how CO2 warms the planet based on a “self-evident belief”?

    That’s a statement of science denialism, not honest skepticism.

  • Ray Ladbury // February 26, 2009 at 2:04 am

    Saltator, the reason Ehrlich was wrong is that he didn’t anticipate that we would learn to eat petroleum (turned into corn, turned into everything else). So what for food do we do when the petroleum runs out?

  • Chris O'Neill // February 26, 2009 at 3:22 am

    Saltator:

    I am not on any propoganda campaign on GW. I know full well that GW is occurring. It is the A bit that I am skeptical about.

    I’ll start believing you’re skeptical when you stop doing unskeptical things like making assertions based on citations that don’t support your assertions.

  • JM // February 28, 2009 at 10:51 am

    “It is the A bit that I am skeptical about. ”

    I thought we’d been through this with our chat about axioms.

    To first order (and probably 2nd and 3rd as well) CO2 is the only greenhouse gas showing any appreciable increase.

    What exactly are you skeptical about?

    1. That CO2 is not increasing?

    2. That an increase in greenhouse gas concentrations will not cause a rise in average temperature?

    3. That a century of burning fossil fuels will not/has not increased CO2 concentrations?

    4. That we have seen a century of temperature increase loosely paralleling our burning of fossil fuels?

    Can you come up with some specifics on the basics or is “skeptical” just a code word for irrational statements along the lines of “nup, don’t wanna pay attention to reality so I’m not gonna, so there!”

  • Saltator // March 3, 2009 at 2:51 am

    A perspective from a bushfire scientist:

    http://www.theaustralian.news.com.au/story/0,25197,25031389-7583,00.html

  • JM // March 3, 2009 at 7:29 pm

    Saltator, I just read the gibberish in your reference and let me offer the following comments.

    The argument appears to be that city greenies prevented bushies from protecting themselves by removing undergrowth and fuel load and are evil, evil, evil.

    Anyone who is at all familiar with the institutional structure of fire management in Australia, particularly Victoria, would know that this is nonsense.

    In rural areas fires are fought by the Country Fire Authority which while funded by the taxpayer is almost totally voluntary in staffing. The volunteers are the homeowners and residents in the area.

    They volunteer to protect their homes and belongings. Even my brother-in-law who is a city greenie with a holiday home in a country town has signed up and does a few weekends training each year. At any time of the 4 weeks summer holiday he spends in his holiday home he will be expected to pull on his yellow jacket at a moments notice, jump on a truck and go and fight a fire - he’s willing to do so and has actually done it a few times over the last 15 years.

    There are no greenies in a bushfire Saltator, it’s a myth.

    Not one of those property owners has the slightest millisecond of time for the ridiculous strawman argument raised in your reference.

    And the pompous old man who wrote it ought to be ashamed of himself, because he is suggesting that my brother-in-law with a substantial property investment accompanied by mortgage is stupid enough to place the shiboleth of some imagined “greenie purity” in front of proper bushfire protection.

    Rubbish. My brother-in-law is not a fool and is not responsible for the small and thankfully not serious fires that went near his house a few weeks ago. We take fire very seriously here.

    The CFA is run by locals and anyone mounting the argument raised by your referenced author as a strawman would be run out of town.

    In any case this thread is about climate change and whether the recent bushfires represent a “canary in a coalmine”, not whether you can get a beer down the pub by sounding off about some mythical evil bureaucrat or “city voter”.

    Get back on topic. What are you skeptical about?

  • Saltator // March 4, 2009 at 2:00 am

    JM

    That article was by a bushfire scientst that has a lot more experience in bushfire research than you, I and just about everyone in this blog put together. He received the Medal of the Order of Australia for his work on bushfire research.

    I will take his knowledge over yours any day.

  • Saltator // March 4, 2009 at 2:16 am

    JM,

    I honestly do not know what you are rabbiting on about. David Packham did not denigrate the CFA or the volunteer bushies. He argued against public policy driven by green politics. That has nothing to do with your brother in law and his personal political leanings.

    I too was a member of a volunteer bushfire brigade until I moved to the city.

    You seem overly sensitive about something here. So much so that you competely missed the point of Packham’s article.

  • Nathan // March 4, 2009 at 4:29 am

    Saltator
    “He argued against public policy driven by green politics. That has nothing to do with your brother in law and his personal political leanings. ”
    This is garbage. What policy in Australia is run by green politics? What green policies were in action?

    This is the conclusion your genius bushfire scientist comes to over the ‘green politics’:
    “The conclusion of these academics is that high intensity fires are good for the environment and that the resulting mudslides after rains are merely localised and serve to redistribute nutrients. The purpose of this failed policy is to secure uninformed city votes. ”

    What a pile of garbage. Has anyone ever claimed this? What science describes extreme bushfires as good? And how could this affect votes in the city? Do you think anyone inn the city has heard of this research? What utter drivel.

  • JM // March 4, 2009 at 7:22 am

    ” So much so that you competely missed the point of Packham’s article.”

    No I haven’t. The greenies vs. the bushies meme is exhausted and hasn’t had traction here for years (other than the occasional rant on talkback radio). Packham’s “point” hasn’t been part of the public debate here following the fires at all.

    The debate (discussion actually) has been focussed on the appropriateness of the “defend or leave early” practice and whether it should be modified in the face of the increasing frequency and intensity of fires - ie. the debate has been resolved and passed the old man by.

    It’s been replaced by a near universal recognition that the nature of fires has changed and we need to change our practices in response - in other words there is a near universal assumption that AGW is probably turning up in the form of drought and fire right now. We’ve left Packham’s view behind.

    What I’m sensitive to is your raising of dead red herrings to avoid a pointed question.

    Which of the elements of AGW are you skeptical about (see my previous comment)?

  • Saltator // March 5, 2009 at 12:52 am

    Nathan and JM

    If you two honestly believe that Green politics has no influence on decision making at all levels of Australian government then you are greater fools than you show yourselves to be.

    and JM. As I pointed out to Tamino some time ago, the philisophical concept of Memetics a moribund one and near to extinction.

    http://cfpm.org/jom-emit/2005/vol9/edmonds_b.html

  • Ray Ladbury // March 5, 2009 at 2:49 pm

    Saltator, at a minimum, Packham’s opinion is merely a political opinion. He certainly provides no evidence to support it.
    Fire does play a role in the ecology of Australia–as it does in most arid environments. We do well to recognize that role and the consequences it has for us when we decide to inhabit such environments. That’s not “green”. It’s not “bush”. It’s realism and a recognition of the tradeoffs between health of the ecosystem and safety of habitation.

  • FredT34 // March 5, 2009 at 11:21 pm

    Saltator, you said on Feb 18 : “Ok. Back to the drawing board. Thanks.”

    So, what results did you obtain ?

  • Nathan // March 6, 2009 at 12:15 am

    So Saltator, you can’t find the reference to the paper on huge fires being good? I guess it must only exist in imagination land then…
    So how big an influence does ‘green politics’ have?

  • JCH // March 6, 2009 at 12:38 am

    In fact, the governmental entity he was bellyaching about has a regulation that requires prescribed burning. They send out notices to people who have failed to do the burning required, and they have a permit process that allows people to burn fuel off.

    I posted links to those sites a long time ago.

    And to be honest, the people in those entities that he calls greens sound far more like what we call homeowners associations - all about development and increasing property values, and little about environmentalism.

    So hopefully the study group will assemble facts, and not just hysterical right-wing rants from editorial pages.

    Those with long memories will remember the california fires during the Clinton administration. Suddenly the Republicans in congress and wingnuts on the radio were experts on controlling fires, and they appeared perfectly confident they would eliminate the source of the California fires - liberal policies.

    Well, 9 years later, California still burns, and as fierce, if not fiercer, than ever.

  • JM // March 6, 2009 at 1:35 am

    Saltator, we’re not talking semiotics, dialetic materialism, cultural discourse or anything else like that - we’re talking physics.

    If you don’t like the word ‘meme’ I’ll withdraw it and offer the following phrases:

    “The false debate of greenies vs. the bushies is exhausted …”

    “The wedge of greenies vs. the bushies is exhausted …”

    or any other version of the same idea you like.

    But that’s not pertinent to this thread.

    You say you don’t doubt the GW bit of AGW, just the ‘A’

    Which element of AGW are you skeptical about?

    Because unless you doubt one of the 4 elements I put forward (or the stronger more precise version that David Benson put forward in the Open Thread #10 (http://tamino.wordpress.com/2009/01/31/open-thread-10/#comment-29306 referencing Levenson and Galasyn) you have no choice but to admit the reality of of AGW.

    Unless you can knock down those arguments you can’t go around proposing unlikely alternatives such as cosmic rays (heaven forbid) or changes in solar irradiance. They’re not necessary if we already know the cause from simple heat balance arguments.

    On the other hand, if you can’t knock down those arguments you are going to have to come up with some Gaia-like natural process that mitigates human activity. And I can’t think of any ever being raised aside from some vague hand waving, or nihilistic sub George Carlin statements along the lines of “we’re stuffed, but the planet will be fine so what’s to worry about?”*

    You say you doubt AGW, can you be substantive and to the point about those doubts please and stop raising red-herrings?

    * Courtesy my geologist brother a couple of years ago - he’s a convert now however.

  • Saltator // March 7, 2009 at 1:48 pm

    FredT34,

    Still working on it. Had to go back to work after leave so it is on the backburner for a bit.

  • Saltator // March 7, 2009 at 1:53 pm

    Nathan,

    “So Saltator, you can’t find the reference to the paper on huge fires being good? I guess it must only exist in imagination land then…
    So how big an influence does ‘green politics’ have?”

    I didn’t look for a reference as it was not I that was making the statement. It was David Packham. I suggest you email him to verify his concerns (you can obtain it via Monash University’s web site).

    As for the influence of green politics. One name:

    Bob Brown.

  • Hank Roberts // March 7, 2009 at 4:54 pm

    Why don’ t you at least tell us what you _think_ you’re talking about, to narrow it down?
    You’re asking us to email Packham and say something like:
    “some guy on a blog, named ‘Saltator’, says Packham or somebody else said something that Saltator hasn’t been able to explain very clearly. Do you deny it and can you prove he’s wrong?”

    So far you’re asking us to play the pony game — deliver a load and promise there’s a pony in it.

    Could you mean this?
    [PDF] ►Peter Attiwül Submission to The Inquiry into the Impact of Public Land Management Practices on …

    P Attiwül - Victoria, 2007 - parliament.vic.gov.au
    … Professor Mark Adams on a book aimed to bring some commonsense to the management
    of fire in Australia’s forests. Attiwill, together with David Packham …
    http://www.parliament.vic.gov.au/enrc/inquiries/bushfires/Submissions/177%20Mr%20Peter%20M%20Attiwell.pdf

    This?
    A coupled atmosphere–fire model: Convective feedback on fire-line dynamics
    TL Clark, MA Jenkins, J Coen, D Packham - Journal of Applied Meteorology, 1996 - ams.allenpress.com
    Cited by 67

    You can look this stuff up as easily as anyone else, and you think you know what you’re looking for.

    Narrow it down a bit past the armwaving.

    Evacuation in wild fires: The Australian experience
    D Packham - Australian Journal of Emergency Management, The - search.informit.com.au
    Cited by 6

  • JCH // March 7, 2009 at 11:42 pm

    Bob Brown is a Senator from Tasmania who opposes clear cutting old-growth forests.

    He embraces burning off fuel loads .

    You don’t have to be Sherlock Holmes to know some cess has escaped the cesspool.

  • Saltator // March 8, 2009 at 4:36 am

    FredT34,

    I have completed my re-analyses and the outcome is the same. The 2000s are not significantly different to any other decade apart from the 1980s.

    I recalculated the anomalies using the 1961-1990 baseline used by BOM. The mean of the corresponding day of each year for the base period was calculated and then subtracted from the value of that day in each year of the dataset to derive the anomaly value.

    The Tukey’s HSD table is below:

    P value adjustment method: holm
    Tukey multiple comparisons of means
    95% family-wise confidence level

    Fit: aov(formula = summer$temp ~ factor(summer$decade))

    $`factor(summer$decade)`
    diff lwr upr p adj
    1960s-1950s 0.148948577 -0.63174309 0.9296402 0.9943322
    1970s-1950s -0.007317073 -0.78822497 0.7735908 1.0000000
    1980s-1950s -0.296344889 -1.07703656 0.4843468 0.8886636
    1990s-1950s 0.180442267 -0.60068227 0.9615668 0.9863140
    2000s-1950s 0.762658764 -0.09761951 1.6229370 0.1163522
    1970s-1960s -0.156265650 -0.93695732 0.6244260 0.9929122
    1980s-1960s -0.445293466 -1.22576884 0.3351819 0.5809119
    1990s-1960s 0.031493690 -0.74941468 0.8124021 0.9999972
    2000s-1960s 0.613710187 -0.24637181 1.4737922 0.3229466
    1980s-1970s -0.289027816 -1.06971948 0.4916638 0.8988311
    1990s-1970s 0.187759340 -0.59336520 0.9688839 0.9836244
    2000s-1970s 0.769975837 -0.09030244 1.6302541 0.1097256
    1990s-1980s 0.476787157 -0.30412122 1.2576955 0.5048205
    2000s-1980s 1.059003654 0.19892165 1.9190857 0.0060212
    2000s-1990s 0.582216497 -0.27825844 1.4426914 0.3844267

    Sorry about the formatting (Tamino really should install TinyMCE to allow formatting). As you can see, the only significant difference is between the 1980s and 2000s (with a P<=0.01).

    The results indicate that the temperatures for Melbourne in the 2000s is not significantly warmer than any other decade apart from the 1980s (which was the decade of the Ash Wednesday fires and others).

    Of course the caveat is that my data stop at December 2006. Therefore, things may change when data for 2007-2009 are added when they become available. Also this is not a trend analysis and trend effects may not yet show for this time scale.

    My R script follows (if Tamino allows it as it may be a bit long) for checking by anyone who wishes.

  • Saltator // March 8, 2009 at 4:37 am

    My R script:

    url = “http://www.mdbc.gov.au/subs/seaci/datafiles/tmax/melbourne”
    mel = read.table(url, header=F, col.names=c(”day”, “dayYear”, “temp”), na.strings = c(”-999.9″, “-999.0″))
    attach(mel)

    yr = (substring(mel$day, 1,4))
    mo = substring(mel$day, 5,6)
    dy = substring(mel$day, 7,8)
    md = substring(mel$day, 5,8)

    yr = as.numeric(yr)
    mo = as.numeric(mo)
    dy = as.numeric(dy)

    mel.data = data.frame(year = yr, month = mo, day = dy, daymo = md, temp = mel$temp, stringsAsFactors = F)

    # remove mel, yr, mo, dy, md
    rm(mel, yr, mo, dy, md)

    # remove missing values
    mel.data = na.omit(mel.data)
    attach(mel.data)

    # generate the baseline anomaly average temp (baseline as per
    # BOM baseline of 1961-1990)
    mel.base = data.frame(year = year[year>1960 & year1960 & year1960 & year1960 & year1960 & year<1991])

    # get mean value for each day of each month
    mel.av = aggregate(mel.base$temp, list(mel.base$daymo), mean)
    # assign names to the two columns of mel.av
    names(mel.av)[1] = “daymo”
    names(mel.av)[2] = “anom.temp”

    # merge mel.data and mel.av
    dat = merge(mel.data, mel.av, by.x = “daymo”, by.y = “daymo”, all = T)

    # remove mel.data
    rm(mel.data)

    # generate a temperature anomaly data.frame
    me.anom = data.frame(year = dat$year, month = dat$month, day = dat$day, temp = dat$temp, anom = dat$temp - dat$anom.temp)

    me.anom = me.anom[sort.list(me.anom$year),]

    # calculate the yyymm variable
    yymm = 0
    yr = me.anom$year
    mm = me.anom$month

    for(i in 1:length(yr)) {if(mm[[i]] == 1) yymm[[i]]<-yr[[i]]+0.04
    else if(mm[[i]] == 2) yymm[[i]]<-yr[[i]]+0.12
    else if(mm[[i]] == 3) yymm[[i]]<-yr[[i]]+0.21
    else if(mm[[i]] == 4) yymm[[i]]<-yr[[i]]+0.29
    else if(mm[[i]] == 5) yymm[[i]]<-yr[[i]]+0.38
    else if(mm[[i]] == 6) yymm[[i]]<-yr[[i]]+0.46
    else if(mm[[i]] == 7) yymm[[i]]<-yr[[i]]+0.54
    else if(mm[[i]] == 8) yymm[[i]]<-yr[[i]]+0.62
    else if(mm[[i]] == 9) yymm[[i]]<-yr[[i]]+0.71
    else if(mm[[i]] == 10) yymm[[i]]<-yr[[i]]+0.79
    else if(mm[[i]] == 11) yymm[[i]]<-yr[[i]]+0.88
    else if(mm[[i]] == 12) yymm[[i]]= 1948 & yr[[i]] <= 1949) decade[[i]] = 1950 & yr[[i]] <= 1959) decade[[i]] = 1960 & yr[[i]] <= 1969) decade[[i]] = 1970 & yr[[i]] <= 1979) decade[[i]] = 1980 & yr[[i]] <= 1989) decade[[i]] = 1990 & yr[[i]] <= 1999) decade[[i]] = 2000 & yr[[i]] <= 2009) decade[[i]] = 38) hot[[i]] <- 1
    else if (melb.anomaly$temp[[i]] < 38) hot[[i]] <- 0
    else (melb.anomaly$temp[[i]] <- 0)}

    melb.decade = data.frame(decade = melb.anomaly$decade[decade != "1940s"], year = melb.anomaly$year[decade != "1940s"], month = melb.anomaly$month[decade != "1940s"], temp = melb.anomaly$anom[decade != "1940s"], hot = hot[decade != "1940s"])
    attach(melb.decade)

    # SUMMER
    # convert summer data to a new data.frame
    summer = data.frame(decade = melb.decade$decade[month == 12 | month == 1 | month == 2], year = melb.decade$year[month == 12 | month == 1 | month == 2], temp = melb.decade$temp[month == 12 | month == 1 | month == 2], hot = melb.decade$hot[month == 12 | month == 1 | month == 2])
    # generate a boxplot of temperature by decade
    boxplot(summer$temp~summer$decade, main = “Melbournce Mean Summer Temperature Anomaly”, xlab = “Decade”, ylab = “Anomaly (degC)”)

    # run analysis of variance on mean summer daytime max temperature
    # by decade
    model.summer<-aov(summer$temp~factor(summer$decade))
    # Plot the Tukey’s HSD data

    quartz()
    plot(TukeyHSD(model.summer))
    # generate a pairwise t.test table for summer
    print(pairwise.t.test(summer$temp,summer$decade))
    print(TukeyHSD(model.summer))

  • Saltator // March 8, 2009 at 6:11 am

    This is a frequency table of hot days (>35C defined by BOM) by decade for Melbourne:

    frequency
    1
    1940s 18 (incomplete only 1948 and 1949)
    1950s 84
    1960s 109
    1970s 63
    1980s 100
    1990s 90
    2000s 80 (incomplete 2007-2009 missing)
    >

  • Saltator // March 8, 2009 at 7:20 am

    The script I posted won’t work. Best to get it from: http://climatebalance.wordpress.com/resources/

    The script is a word document called Melbourne Temperature Anomaly Script.

    Copy and paste it into R.

    If using R for windows change quartz() to windows()

  • FredT34 // March 10, 2009 at 9:36 pm

    Saltator, thanks for posting this. Unfortunately, I am not capable of reading your code or checking it (I hope others will) - but I appreciate your posting : it proves that you’re going on with things… I must admit I suspected you as being some “I read this somewhere but don’t have the link right now” guy.

    The only thing I can read in your table above is : the frequency of “hot days” may well go above 110 days for the 200x decade… as 3 years out of 10 are missing, and 2009 was terrrible.

    The other thing is : data, data, data… Without data (good data of course) we can’t calculate trends nor elaborate or test models. For those things, the loss of Cryosat and OCO are very bad…

    Thanks,

    Fred

  • Saltator // March 11, 2009 at 3:34 am

    FredT34

    I have since added 9 other locations to the dataset. The result changes significantly. When I add locations from around Victoria, the ANOVA reveals that, in fact, the 2000s is significantly different from the other decades.

    I don’t think that we can say this is AGW per se but that it may be an artifact of 10 years of drought dominated by El Ninos that have elevated temperatures during the 2000s.

    Others may differ on that issue though.

    I think that 2009 adds a number of days to the hot days for the 2000s and 2007 and 2008 may add enough for the frequency of hot days in the 2000s to exceed those of the other decades.

  • FredT34 // March 12, 2009 at 12:40 am

    Saltator : Thanks for this research. I feel this reinforces my opinion that “an average is more significant than any single-data - which can always be cherry-picked”. And, trend on averages is more significant than this-or-that data - it’s always easy for a debater to find meaningful examples - but non-significant… The problem with averages is that they’re much less spectacular…

    About El Nino… of course it was powerful at some times during that 2000 period. But one can think it also was during previous periods - so I’d tend to think that GW, for some part, exacerbates El Nino (and possibly La Nina, and all extreme phenomenons), which in turn perhaps contributed to this heat wave.

    Well we had a big heat wave here in France in 2003, and I’m quite anxious to see it coming back - as Australian will anxiously watch heat and fire forecasts in next years…

Leave a Comment