An Assessment Of The Predictions Of The Commercial Company Global Weather Oscillations Inc And The Onset Of Global Cooling.

[johnny1972]

Iceberg

Nice post but please look at the following:

Test 2 Phase 1 global cooling: There has been a 0.1 cooling in both the GISS and HADCRU3 data sets compared to recent years. However the measured cooling does not support the prediction. MSU does support the prediction and would need to see a continuation of the average temperature seen so far this year for the prediction to be fulfilled.

It should be noted that all three temperature series have followed a warming trend so far this year though and that if this were to continue all would result in a failure of the predictions measured.

Maybe i have missed something but looking at the table on the download and comparison to recent years the averages with all 3 data sets are clearly downward trending cooler than 2006 and 2007. Ok it maybe doesnt fit exactly with the prediction but cooling the world is. Weather forecasting isnt an exact science so yes i would agree further assessment of the data sets over the next few years would be required to see if the cooling trend continues.

[BLAST FROM THE PAST]

to clear up any ambiguity as what GWO did or did not say, quoting from the link given by Paul

pulled out of its normal location in October and November, setting the stage for a moderate to strong El Niño to form in December.

thus IF it has not become moderate to strong by 31 December then that part of the prediction can be said to be wrong, equally if it is at least moderate then GWO can feel his prediction for that part is on track?

A difficult one to decide and the term 'form' is ambiguous. Now if El Nino conditions form [start] in Dec and continue on to become Moderate to strong in say March then that El Nino 'formed' in December. Then again it can be read as some are suggesting. The point of saying form rather than 'by' suggests that we only need to see El Nino conditions in December BUT they then must grow to moderate to strong conditions.

BFTP

[Roger J Smith]

Here's a map reference that might be of interest for those looking out for the El Nino ... current and days one to four of the region south of the Peruvian coast at any rate, would expect this to show any unusual weakening of the subtropical high in the South Pacific. It doesn't look exceptionally strong right now but it's fairly stable. I checked today's temps along the Peruvian coast, the usual near 20 C sort of thing which for that latitude shows that there is cold water and low cloud in the picture.

Here's the link:

http://www.smn.gov.ar/?mod=dpd&id=1

[Roger J Smith]

After thinking about the validation proposal, I wanted to make the following comments.

When David speaks about a general cooling to become evident in the time frame 2008-09, I think a lot of us who are interested in this subject may be more influenced by events in the mid-latitudes than the subarctic. From my point of view, I will explain why. It is clear from the data in the three global inventories being used that we have entered a period of slightly higher temperatures that are maintaining themselves at least in these indices, and I suspect a lot of this is due to the robust nature of the SST anomalies in the subarctic and especially the Svalbard to Novaya Zemlya region. If that region contributes about 10% of the data base and is staying 2-4 C warmer than its long-term averages, then this will skew the rest of the hemispheric or global data, but this would likely be the last region or sub-region to see a cooling, as the climate is entirely maritime in nature and profoundly influenced by small changes in ice margin or ocean currents.

So I would be fairly impressed with David's forecast even if there were slight positive errors in these tracking inventories, as long as there were some sizeable negative anomalies in the mid-latitudes on land where people actually live.

All these things are linked, but I am not sure if the physics of the atmosphere can sustain a sharp cooling trend over the next 3-6 months for example, before we see some recovery of the ice margins around Svalbard to Franz Josef Land in particular. The atmosphere is possibly entering an unstable phase here, with cooling influences trying to overcome residual warming effects, and so what's probably more important to both understanding and credibility of method is whether or not there is a general decrease in hemispheric if not global temperature over the next 3-6 years. I would consider it a bit premature to take this winter and say, well it has remained globally warmer than base-line values, so this theory must be thrown out, when in fact it may be predicting processes that are starting up -- this is one of the reasons I was cautioning against using non-statistical terms like "100% correlation" because in a breakthrough theory of any kind in this field, a correlation in the order of .6 or .7 would be astonishingly good, yet only explain 36 to 49 per cent of the variance (at least that's one of three things I recall from my university days which are a good half century removed almost).

I'm not trying to subvert the process, but I wouldn't want the process to subvert the research either. David's research is valuable on a much larger time scale than our attention spans here on NW, this is one of the frustrations of doing this kind of work as I well know, that you are dealing in longer time scales than people can sustain an interest, and by the time some really valid concept on the 20-40 year time scale might be possible to verify, people have moved on, to another interest, or perhaps another world.

Before this global warming theory became all that much the rage, people were making forecasts of future trends and some were quite accurate about the 1990s and early 2000 decade being relatively warm, as I dimly recall. These were based mostly on natural variations. But whether anyone remembers making or criticizing these forecasts in the 1970s or early 1980s is not that easy to say.

I would also like to establish from the outset that validations of one theory should not be used to draw conclusions about another theory's validity. This works both ways in this instance, David and I have different research programs and we are just getting into the process of understanding where the other is "at" if you see what I mean. If my winter forecast busts, that's no reflection on David's modelling, and if his El Nino forecast is seen to be off target, that has nothing to do with my technique. I'm still hopeful that these different research ideas can be blended together but I haven't made any real progress on that yet, too many other things going on. I see some overlaps already, and some points of difference between our projections. And with the subject matter being as complex as it is, I readily suspect that there could be other variables of the same kind that we both need to add to our work to get everything "up and running." I think there's more work left to be done than has been done so far in this developing research field.

[Iceberg]

A difficult one to decide and the term 'form' is ambiguous. Now if El Nino conditions form [start] in Dec and continue on to become Moderate to strong in say March then that El Nino 'formed' in December. Then again it can be read as some are suggesting. The point of saying form rather than 'by' suggests that we only need to see El Nino conditions in December BUT they then must grow to moderate to strong conditions.

BFTP

Thats all well and good but on his press release 3 on his web site it says "by".

Mr. Dilley says the PFM gravitational cycle that controls the formation of an El Niño peaks approximately every 4 years, and it will again peak during this September into January. As this occurs, the tropical South Pacific Ocean and atmospheric winds will respond rapidly and cause the formation of a moderate to strong El Niño by Christmas.

I think it's very clear what he means.

Iceberg

Nice post but please look at the following:

Test 2 Phase 1 global cooling: There has been a 0.1 cooling in both the GISS and HADCRU3 data sets compared to recent years. However the measured cooling does not support the prediction. MSU does support the prediction and would need to see a continuation of the average temperature seen so far this year for the prediction to be fulfilled.

It should be noted that all three temperature series have followed a warming trend so far this year though and that if this were to continue all would result in a failure of the predictions measured.

Maybe i have missed something but looking at the table on the download and comparison to recent years the averages with all 3 data sets are clearly downward trending cooler than 2006 and 2007. Ok it maybe doesnt fit exactly with the prediction but cooling the world is. Weather forecasting isnt an exact science so yes i would agree further assessment of the data sets over the next few years would be required to see if the cooling trend continues.

Yes 2008 so far is cooler than 2007 or 2006. Your right it doesn't fit exactly as the cooling trend needs to be multiplied by a factor of 5 in order to met the prediction. WRT the temperature prediction I think a monthly update with a half way point taken at the end of 2008 would be fair.

At the moment we've entered a race travelled a little way down the track and stalled.We could get going though and finish it but the stalling can't be seen as positive.

[Iceberg]

Roger, I've said on several occasions that I don't want this thread to prove or disprove GWO's theories, it's purely about assessing the accuracy of his predictions.

I have an image of GWO graphs that show a direct correlation and predictable timelapse between the PFM and temperature as well as ENSO phase changes. The claim is that has shown accurately every ENSO event and all temperature changes globally.

I am more than happy for GWO to clarify what he means by the predictions to enable the assessing to be done as fairly as possible.

[The PIT]

Err but if you're assessing you're judging the quality of the predictions so basically you are trying too prove or disprove the predictions. Surely then if the predictions are wrong then the theory is hanging somewhat by a thread????

[Iceberg]

That would be depend on whether or not the predictions were made accurately from the theory. Does that make sense.? Also the theory as Roger points out could be 60% accurate rather than the near 100% claimed currently. It would be upto GWO to respond though better than me.

[Arnost]

That would be depend on whether or not the predictions were made accurately from the theory.

A very interesting point.

When the predictions were made - I'm guessing some six months ago or so - the state of the Pacific equatorial sub-surface was such that a formation of a Nino by year end was odds-on. The models also supported this view. This coupled with a (admitedly very weak) correlation between El Ninos and the start of solar-cycles would have made it logical to project a Nino - and even a strong one - in December.

In my opinion, it is now extremely unlikley that a Nino will form in the next two months. The current sub-surface, the SOI, the OLR, the GLAAM, and the Trade Winds all support the story that the ocean and the atmosphere is in a Nina state. And the sea surface is slowly coming into line as well (Nino 3.4 was already at -0.5C a week or two back). The persistence of ENSO state across the end of year and beyond is a known feature. Whether a full blown Nina delvelops - who knows, but I would suggest that there is now no chance of a warm-neutral ENSO state, let alone a Nino of any kind, well into the new year.

Now - trying to throw out a lifeline... about a month ago there was a WWB (westerly wind burst) in the west Pacific that generated a Kelvin wave that's now pushing a sub-surface warm pulse slowly across the Pacific. And it appears to be reasonably strong... It has a (admitedly) small potential of killing off the developing Nina. And the sun's startig to be a bit more active. Contrary to what the current models - where the latest (blue) runs suggest a moderate/strong Nina early next year - there may be rapid turnaround to a Nino state - and we may have the strongest Nino of this decade in 09/10. This is worthwhile watching, but IMHO a long shot at the moment.

So - did Mr Dilley get swayed by the consensus on the future state of the ENSO when making his prediction or was he true to his model/theory?

Edited October 21, 2008 by Arnost

[jethro]

That would be depend on whether or not the predictions were made accurately from the theory. Does that make sense.? Also the theory as Roger points out could be 60% accurate rather than the near 100% claimed currently. It would be upto GWO to respond though better than me.

Surely when the IPCC and their theory/predictions are tested against this benchmark they fail? Where does that leave their theory? To date, none of their predictions have been accurate or come true.

[Paul]

I think that's a subject for another thread though isn't it? This is to assess GWO's theory, perhaps someone could do something similar with the IPCC's work and indeed any other theorists out there and open a separate thread for that assessment?

[Red Raven]

I think that's a subject for another thread though isn't it? This is to assess GWO's theory, perhaps someone could do something similar with the IPCC's work and indeed any other theorists out there and open a separate thread for that assessment?

I agree it is something for another thread but it is important in any theory testing that it is the actual theory that is tested and not the popular press interpretations of it that go under scrutiny.

[jethro]

I think that's a subject for another thread though isn't it? This is to assess GWO's theory, perhaps someone could do something similar with the IPCC's work and indeed any other theorists out there and open a separate thread for that assessment?

Fair comment; I am struggling to see how GWO is going to be judged or measured though, other than the El Nino, I'd say it's too early to tell. If the Nino doesn't verify, does this invalidate, some, all or none of David's theory? How do we judge/measure this?

Genuine question btw.

[Paul]

As I understand it, Iceberg is not looking to prove or disprove the theory as a whole, just validate certain specifics within the theory. What conclusions people decide to draw from that validation is probably a whole new discussion too (lol I sound like a stuck record).

I'd think though, it's validation such as this which really is the only way forward when it comes to the climate change debate, people can argue about theories, opinions and interpretations until the cows come home, but when all is said and done, assessing the facts surely has to be the answer?

[jethro]

As I understand it, Iceberg is not looking to prove or disprove the theory as a whole, just validate certain specifics within the theory. What conclusions people decide to draw from that validation is probably a whole new discussion too (lol I sound like a stuck record).

I'd think though, it's validation such as this which really is the only way forward when it comes to the climate change debate, people can argue about theories, opinions and interpretations until the cows come home, but when all is said and done, assessing the facts surely has to be the answer?

Absolutely, assessing facts is essential, it's just, as far as I can tell the only bit which can be assessed is the El Nino - this all seems a tad premature to me. I expect I'm just being my usual picky, pedantic self so will bugger off, wait and see how this pans out.

[Gray-Wolf]

It's also having the 'fact's' accepted as 'fact's' (or so it would seem).

No matter which 'fact's' are presented someone will always question their validity by stretching the envelope of the subject matter to the point where the 'fact' can be viewed as unsubstantiated.

The years I've spent in open discussion with what I have held as empirical 'fact' has, by necessity, led me to this conclusion.It further leaves me suspecting that in any 'debate' on line is more about personal 'viewpoints' (and an unwillingness to be 'converted') than actually moving forward our collective understanding.

Any person who puts forward a personal hypothesis is likely to fall prey not only to it's possible failure but also the failure of folk to keep an open mind whilst awaiting the conclusions.....

Ho Hum.

[Roger J Smith]

Just to clarify somewhat my earlier remarks about correlation, I think GWO was saying "100% correlation" as a sort of expression of enthusiasm for the general similarity between his PFM index and temperature reconstructions over a long time frame. I don't think he meant "100% correlation" the way a statistician would mean it by saying "correlation 1.0" and I had a look at the graphs involved and tried to devise a way to test the correlation, but I was unable to work out how I would do that from the complexity of the theory which shows how two separate PFM cycles govern the temperature signal in the reconstruction. If you see what I'm saying, I generally agreed with David that the correlation looked fairly high and that this seemed like a real process but because it is being driven by two sets of signals, I am still working on some sort of formula that links the two signals so that I have a composite PFM predictor to correlate with temperature.

Even so, this technique, whether it has a correlation of .5 or higher, applies to long periods of temperature data so if it were let's say 0.7, this would not necessarily inspire confidence that the technique would immediately yield that kind of correlation on a month to month scale where the reconstruction has smoothed out all the variations.

That's what I'm getting at here, the statements being validated are over a fairly short time scale whereas the most compelling demonstrated skill of the model appears to be at time scales quite a bit longer than year to year. Even so, there are indications of correlation on a year to year or seasonal basis. My general conclusion remains the same, this is a pathway to understanding what is happening in the atmosphere. Whether this first round of forecasts is the best illustration of that potential or not, the longer-term correlation establishes that this would be very worthwhile for a lot more development and research. So this validation is just going to tell us how this one effort is going, but even if it were to prove totally inaccurate I wouldn't want to overlook the obvious potential of the method on the longer time scale, and sometimes these first efforts to shape a theory to shorter time frames are just lacking in detail or complexity.

Sooner or later I think this general approach (which I call astro-climatology) is going to start working more and more predictably and then it will become obvious to all that this is the right paradigm for monthly, seasonal and annual forecasting. I wish we could get more people up to speed on what is already known and then this would naturally go faster with more minds working on complex problems. Nevertheless, I feel that we are getting closer and that this sort of breakthrough period for the field may be within sight now, let's say in the decade ahead to give a rough estimate. I probably have ten to fifteen more years left in the tank and huge amounts of data already analyzed waiting for the next round of modelling to begin (just getting into that now after finishing about two or three years of analysis). So in my own case I won't really know how much progress has been made until perhaps 2010 with a year or two of forecasts based on the latest research in the bag. I would say then, expect perhaps some developments in this area of research fairly soon, but we are probably at that stage where rockets are being designed and launched and a certain number of them go off course or blow up on the launching pad. That's bound to be the situation with any developing global climate model.

[BLAST FROM THE PAST]

Hi Roger

As I see it there is going to be a need for some neartime accurate 'hits' that will make folk really jump on board. There is movement in the right direction but a direc El nino hit for David and a direct winter 'hit' for yourself will propel enthusiasm.

Me I see it already but hey I'm biased aren't I?

BFTP

[Cycles]

I do recall a rapid (and unexpected) El-Nino/ La-Nina flip flop before the 2006 hurricane season which left the NOAA predictions of storm numbers in tatters (along with many of the disturbances that tried to form due to conditions 'not condusive' to 'cane formation).

The strongest El-Nino 'this century' would'nt need to be that strong seeing as we're only 8 years into it.......

Though still officially 'Neutral' conditions across the pacific the conditions are pretty mixed and, as quickly as the cold water upwelling squished the last Nino into neutral, I imagine the areas of subsurface warm wouldn't take much to impose 'nino conditions over just a matter of weeks.

As you know I believe we are in 'strange times' so nothing would surprise me! B)

Forecasters of the El Nino have certainly been caught off guard a few times, they can form very rapidly. A very strong segment of the PFM cycle will occur between October to January, this is expected to displace the tropical high pressure system, stop upwelling in the central South Pacific Ocean and cause rapid increases in sea surface temperatures. It is a waiting game right now, when the high shifts so will the winds causing upwelling.

Regards

David

[Iceberg]

Forecasters of the El Nino have certainly been caught off guard a few times, they can form very rapidly. A very strong segment of the PFM cycle will occur between October to January, this is expected to displace the tropical high pressure system, stop upwelling in the central South Pacific Ocean and cause rapid increases in sea surface temperatures. It is a waiting game right now, when the high shifts so will the winds causing upwelling.

Regards

David

GWO are you saying that the ENSO moderate/strong EL Nino phase might not happen until January now ?.

If so I think it's a valid change to your prediction and extending it by a month won't invalidate the prediction in anyway.

I am happy to keep monitoring the ENSO TEST1 for an extra month.

Forgot to add could you provide a rough lat and long of the high pressure cell and where you are expecting it to move to I will happily add this as well.

[Cycles]

Below is an attachment showing the ONI issued by NOAA for the period 1950 to Oct 2008. Note El Nino events normally form once the ONI increases to zero, it is currently at zero.

oct2008oni.pdf

[BLAST FROM THE PAST]

Intersting to see the jet plunging south in the US bringing November like cold to large swathe of the US.

BFTP

[Cycles]

GWO are you saying that the ENSO moderate/strong EL Nino phase might not happen until January now ?.

If so I think it's a valid change to your prediction and extending it by a month won't invalidate the prediction in anyway.

I am happy to keep monitoring the ENSO TEST1 for an extra month.

Iceberg....by all means, keep the ENSO TEST1 going into January. NOAA uses a 3 month average or more (will get correct definition for El Nino) to determine if an El Nino has formed. Knowing the mechanism controlling ENSO formation, my forecast is for the beginning of an El Nino and thus bye-passes the 3-5 month averaging for determination.

[Iceberg]

Your forecast is for a moderate to strong El Nino to have formed isn't it, as per PR3 on your website ?.

Yes the pdf you posted is a graphical representation of the same data source I used. It classifies anything greater than a 0.5 as an EL Nino.

If your forecast isn't calling for a moderate to strong EL Nino by Jan, when do you think it will be ?.

[Cycles]

Your forecast is for a moderate to strong El Nino to have formed isn't it, as per PR3 on your website ?.

Yes the pdf you posted is a graphical representation of the same data source I used. It classifies anything greater than a 0.5 as an EL Nino.

If your forecast isn't calling for a moderate to strong EL Nino by Jan, when do you think it will be ?.

It is likely in the infancy of formation now with the ONI reaching zero, it should reach +.05 during December. I just cannot pin the actual date for you Iceberg. The mature moderate phase would likely follow the .05.