Data Comparisons

VG predicts UAH June anomaly will go negative.

28 Comments

Like a budding Cassandra, VG has been predicting the upcoming UAH anomaly will be negative. As early as June 11, he posted this:

Be prepared for a 0 or below for June 2009 http://discover.itsc.uah.edu/amsutemps/ channel 5 There is no El Nino…. yet.

I replied,

VG– Don’t count your chickens before they are hatched. We’ll need at least a few days below 0 to get the monthly averages below zero. Still…. that UAH measurement did cool down smartly over the past few days.

Nevertheless, I have been curious, so visited Daily Earth Temperatures from Satellites and took a peak. Then another. Today, I thought “What the heck! I’ll download the data and process it to see what the ‘Channel 5’ anomaly is”. To create the ‘Channel 5 anomaly’ I subtracted the 20 year average from the 2009 measurements. I plotted these and had a look:

Figure 1: Channel 5 daily anomalies.
Figure 1: Channel 5 daily anomalies.

The average channel 5 anomaly is negative based on June 1-21. That’s 2/3rds of the month.

Given how the daily anomalies jump around, I still think VG shouldn’t count his chickens before they hatch. After all, the first two weeks in Chicago were the record cold start of June. But I mowed the lawn yesterday; I took a walk today. Whoo, boy. It’s hot out now!

I’m betting the final average for June will not hit a record low! If Chicago can warm up, I bet the troposphere could warm up over the next week and the negative channel 5 reading will vanish.

Also, remember that the UAH measurement is not computed by taking the average of the daily measurements from channel 5. Roy Spencer warns,

The progress of daily temperatures should only be used as a rough guide for how the current month is shaping up because they come from the AMSU instrument on the NOAA-15 satellite, which has a substantial diurnal drift in the local time of the orbit. Our ‘official’ results presented above, in contrast, are from AMSU on NASA’s Aqua satellite, which carries extra fuel to keep it in a stable orbit. Therefore, there is no diurnal drift adjustment needed in our official product.

So, yes. Channel 5 indicates a cool troposphere. (Interestingly, since 1999, Channel 5 has tended to indicate a cool troposphere in June. However, since we don’t know the diurnal drift correction for channel 5 for this June or even June in general, we can’t really know what that means for the full June 2009 reading.

Maybe VG will be right. Or not. Knowing Roy, he’ll tell us sometime before July 4.

28 thoughts on “VG predicts UAH June anomaly will go negative.”

  2. Anyway why are you worried? You only use Hadcrut, Giss etc I believe? (Stand to be corrected if wrong) They are never negative… LOL

  4. VG–

    ..whats with the Chicago story

    Think global, act local? 🙂

    I show all the temperature metrics. But I compare projections of surface tempartures to surface temperatures. UAH went negative last year. Might happen again. If it does, I’ll try to remember to mention you told us so first.

  5. Hi Lucia This whole thing came about due to David Archibalds prediction that May 09 would hit – 0.4c. I been very impressed so far with his solar predictions say as compared to the really “official geniuses from NASA” who have been way off everytime with sunspot predictions etc.. I got a bit angry when someone critized him (DA) for not getting it right for May at another blog. The point being that there appears to be a decline towards normal or below lately so in this case he may have been slightly off.. but as far as I can see not a bad guess again!

  6. VG,
    I don’t follow Archibald’s predictions. I try to avoid specific predictions of the future. It’s much easier to describe the past.

    You may turn out to be right on UAH for June. So, you might as well get credit for it if you are.

  7. Just to show fairness/non bias.. I posted similar links SST pics some time ago (8 months ago?) showing that oceans seems to be cooling or had vast “areas of cooling”. The same pics today (June 22nd) suggest more “warming areas”
    http://www.osdpd.noaa.gov/PSB/EPS/SST/data/anomnight.6.22.2009.gif
    However the UNYSIS images today ( June 23rd) seems to show more “cooling areas”
    http://weather.unisys.com/surface/sst_anom.html
    Maybe me is color blind LOL. Does anyone know if the same satellite pictures are used?

  8. Here in NZ this has been the earliest, longest, coldest start to winter for at least the last decade. There will be no surprise at all here if the anomaly goes negative. Roll on global warming we say.

  9. VG– We’re entering El Nino. See the red line in the equatorial pacific?

    What are the UNYSIS images relative to? Yesterday? Last month?

  10. I see your cold NZ winter and raise you 100 (37.7 to weirdos) yesterday in Tennessee. Meh, not that unusual for this time of year (though July and August are when most triple digit days hit). Seems worse because my AC went out yesterday morning…

  12. Boris– I was wondering if you were sweltering! We had a conference call with the guys down in Cham-bana (i.e. Champaign-Urbana), one in Florida, one in Maryland and one in New Hampshire monday and another yesterday. Monday morning the Illinois group told the Floridian it had been cool and rainy and one of the guy’s had a father-in-law who couldn’t plant his beans yet because the fields were too soggy. Tuesday we reported that front sitting over Louisiana moved north and we are now sitting in a hot sauna.

    According to Skilling (local weatherman) two fronts have been oscillating over Illinois. The hot front has been over Louisiana/Memphis/Cairo etc. The cold one is north in Wisc. We’ve had downpours as it oscillates north and south. Right now, Chicago is on the Lousianna/Memphis/Cairo side of the front. It’s sunny, but really hot and humid.

    But I’m at least further north than you are. I can tell you, I’m glad I’m not in Cairo Illinois. Memphis or Louisiana would be much worse.

    With all this rain, I will need to mow the lawn again. (I mowed Monday. I mowed Saturday. )

  13. Checked today – Ch5 merrily on it’s way back up.

    By the way Lucia. did you get my e-mail? It just shot off into the ether never to be seen again. Maybe you just chose to ignore it – that’s OK. Just wanted to check to see if your e-mail thingy works. :]

    Love your edit thingy – very nice for us keyboard challenged types.

  17. I am puzzled by recent data. Uah daily does suggest a dip down in temps for June. SST anomlies such as at Unisys suggest a fairly major increase. In particular NCEP release a weekly ENSO update which compares SST against 4 weeks ago and recent updates have been heavily dominated by warming anomalies. See page 8 of presentation at http://www.cpc.noaa.gov/products/analysis_monitoring/lanina/enso_evolution-status-fcsts-web.pdf

    How to reconcile? Maybe the ocean is heating up and land cooling? But NCEP reanalysis data seems to show just as much warmth on land as Ocean for last 30 days.

    So maybe one of these sources is in significant error.

    Or maybe as a warmer I’m suffering confirmation bias when looking at the SST maps and overestimating the red and discounting the blue.

    Or maybe it just representative of the expected differences in satellite sensors measuring SSTs and UAH.

    I also suspect that Uah daily is following a semi regular up and down cycle possibly influenced by the MJO, we should be at the bottom of this cycle now and going up again in the next couple of weeks. I would guess this would be enough to push the channel 5 anomaly just into positive, but still the lowest for at least a few months.

    Its interesting that the dip in Uah from 2007 to 2008 seemed to follow fairly closely a transition from el nino to peak la nina, with peak nino 3.4 in Dec 06, peak UAh in Jan 07, and bottom UAh and nino 3.4 both in Jan 2008. Whereas the current Uah dip started in March with a bottom either now or yet to come, and the closest corresponding ENSO movement is probably the reduction from nino 3.4 peak in Aug 08 to bottom in Jan 09.

  18. I’ve been trying for ages to get a half-decent regression from UAH daily to Hadley/GISS/UAH/RSS monthly numbers. I’ve stuck my neck out a few times to make predictions, and “it has been a humbling experience”. Going from the raw average of the daily UAH numbers, I get no skill at all. The only reasonable skill seems to be when I…

    * do a raw projection for each of Hadley/GISS/RSS and note the slopes of the trendlines. This is *NOT* a forecast.

    * wait for RSS results to come in. Calculate
    delta(RSS) = RSS(projected) – RSS(actual)

    * Hadley(forecast) = Hadley(projected) + (delta(RSS) * slope(Hadley) / slope(RSS))

    * GISS(forecast) = GISS(projected) + (delta(RSS) * slope(GISS) / slope(RSS))

    Effectively, I wait for RSS to provide an “origin” for Hadley and GISS. For some reason, UAH monthly has the worst overall agreement with UAH daily.

  21. So the big question is… where do we find a useful daily global temeperature dataset that’s updated daily near-real-time?

  22. Walter–
    That’s interesting. But I’m not sure it means much other than you can’t predict much using channel 5 daily numbers. Alan Wilkinson explained why (as Roy has doe before him.)

    I’ve looked at correlations between the various agencies from time to time (mostly because the uncorrelated part give the lower bound on the uncertainty due to all factors.) It’s not clear you can learn much by figuring out which agency product is least correlated from the others. Some of the correlation is due to the fact that the agencies draw from a collection of underlying temperature products, picking some subset. So, suppose that there are 8 product measuring 3 component of the temperature: (a,b, c,d,e, f,g,). If an agency A picks (a,c,f) agency B picks (a, c,g) and agency C picks (b,e,g).

    Now, lets do a “what if”. Assume instead of being instruments, all “a-g” do is report random noise. What will we find:

    Agency A and B’s product will be strongly correlated with each other because both used “a” and “c”. But they will differ because f is different from g. Agencies “C” will show some correlation with “B” because both use “g”, but otherwise differ. Agency “A”and “C” will exhibit no correlation at all.

    So, if “B” and “C” always reported before “A”, and you tried to “B” or “C” to predict A, you would find that “C” told you nothing about “A” and “B” told you a lot.

    But this would have little to do with the relative accuracy of any instrument!

  24. Deep Climate–
    No way it’s going to be -0.4C in June! I don’t see anyone suggested that here. (Has anyone anywhere actually predicted anysuch thing?)

    It will be interesting to see if UAH does happen to go negative at all. Channel 5 was at -0.004C for June. Yesterday’s specific value was above zero. Unless Mars attacks between now and June 30, there is no way its June average is going to hit -0.4C! The graph in the figure above shows that since 1999, Channel 5 has tended to be low in June.

    It’s interesting to see that the Alberta boys in your post share Rahmstorf’s difficulty with oscillating trends at end points. Hadley fixed theirs by admitting they don’t know how to smooth or fit trends in a way that gets the trend correct at the endpoint. Do you think Rahmstorf and the Alberta boys will ever learn?

  25. Looks like it could be a tiny bit over 0 or not… LOL. Still the trend seems to be down…. A look at COLA shows large areas of ASIA, South America and AFRICA in a cool mode.. W.Europe Hot mode. etc…

  26. Deep continues to subtly suggest that something is wrong with UAH (innuendo…) but I have not seen him attempt to address published-as opposed to blog-arguments favoring UAH over RSS:
    http://www.agu.org/pubs/crossref/2008/2007JD008864.shtml

    I can’t understand this tribe mentality whereby UAH is universally agreed by the warmers to be “bad”…presumably because it is maintained by two of the “enemy”…but that is not a scientific reason. Nor is less agreement with the models, which data are supposed to be used to test. I really don’t get this…

  27. Walter Dnes (Comment#15208)

    So the big question is… where do we find a useful daily global temeperature dataset that’s updated daily near-real-time?

    Wish I knew. The data are obviously out there, but the nearest I’ve come are these maps from noaa. The problem is the scale is really coarse (white areas are 0 +/- 2.5 deg).

    http://www.cdc.noaa.gov/map/images/rnl/sfctmpmer_01a.rnl.html

    Anyway, it’s fun to check them periodically. Also, you can change the time frame by substituting other numbers for the 01 in the url address. So far, I know of maps for 01, 07, 30, 90, 180, and 365 for maps that cover one day, one week, one month, etc.

  28. John M (Comment#15409) I think it should be obvious why such a data set doesn’t exist. At all.

    See, the data has to go through a lot of processing to correct for various sources of error. That takes time. I figure at least a day.

    Now, maybe a little less than near real time…

Comments are closed.