Those are pretty close to the number that I came up with when looking at the data. I came up with about a 62-63 year cycle. This is the reason that I have been saying that we will not have significant warming for another 15-20 years.

Now there is still a question if we are even talking about the same thing, though.

http://en.wikipedia.org/wiki/Pacific_dec...

http://www.nwfsc.noaa.gov/research/divis...

Note how the PDO is talked about as a climate index based upon the temp of the North Pacific.

What you (and I ) are referring to is the overall cyclical pattern seen in the global temperature data. I have called this the PDO simply because I have no other word to describe it AND it has the same length of cycle as the PDO. But the PDO and what we are talking about need not be the same thing. The actual PDO is NOT determined by the overall global temperature. So it may be that what we are seeing in the global temperature is caused by the PDO but off cycle with the PDO OR it may be that the PDO is just coincidentally on the same length of cycle, but only has minimal effect on the global temp. I personally think the former is true.

What is clear though, is that the way they measure the PDO by the Pacific surface temps does not exactly fall in line with the maximums and minimums in the seeming cycles seen in the global temp data.

As a statistician that has worked with cyclical data before, however, I suggest caution in interpretation. Realistically, we only have enough data to see about 2 to 2.5 cycles of a 60-65 year cyclical pattern. This hits below the minimum of what you want to see. In dealing with cyclical data it is generally good to have enough data to cover at least 3 cycles. So while I think the cycles valid, I would certainly not go so far as to say they are definitely occurring.

Edit:

Also when you think of the PDO as adding or taking heat away from the Pacific ocean, it would make sense that the global temps would slightly trail the PDO in effect. But once again it is difficult to talk with much certainty while having less than 3 cycles worth of data.

Jeff M gives a good answer. I'll just add another link or two.

You can see a plot of it here:

http://www.esrl.noaa.gov/psd/cgi-bin/dat...

and you can make your own plot of it and lots of other climate indices at this page above it:

http://www.esrl.noaa.gov/psd/data/climat...

You can also download the time series data there if you like, to do your own fancy analysis on it. It's not a beautiful sine wave, that's for sure. We don't have that long a record of it, and we're not exactly sure why it switches phases. Some say a big El Nino can do it (it's sort-of a "Super El Nino" itself)

I don't like to get too wrapped up in it. Clearly the curve is jittery enough that even in a low or high phase it may spend some time in the opposite.

http://jisao.washington.edu/pdo/

http://www.nwfsc.noaa.gov/research/divis...

The PDO switched phases, from the looks of it, in about 1948 from warm to cold, 1977 from cold to warm and 1998 from warm to cold. Since 2000 there seems to have been minor fluctuations between warm and cold. Bringing this further into a discussion about how ENSO is related to PDO we get the following:

http://cses.washington.edu/cig/fpt/water...

I think the next phase shift is somewhere in the vicinity of 10-20 years or so.

You have good answers from Jeff and Pegminer, however I do not like referring to it as cold or warm, it is more like heat being added to the oceans (negative phase) or heat being extracted from the oceans (positive phase)

It is a big mystery, we know about it, and we know it cycles, but we do not understand what causes it.

Considering the graph of global average temperatures;

http://woodfortrees.org/plot/hadcrut3gl/mean:60/plot/hadcrut3gl/from:1880/to:1910/trend/plot/hadcrut3gl/from:1911/to:1939/trend/plot/hadcrut3gl/from:1940/to:1972/trend/plot/hadcrut3gl/from:1973/to:2004/trend/plot/hadcrut3gl/from:2005/trend/plot/hadcrut3gl/from:1850/to:1879/trend

it can be seen that, as well as a general warming trend, there is a clear cyclic component, (which most people would agree was the PDO) and some noise (El Nino, volcanoes etc)

What I am interested in is the times that that the PDO is considered to have changed phase from positive to negative and vice versa. From my link I would say around 1880, 1911, 1940. 1973 and 2005.

But that's just me playing around on the woodfortrees site! Loads of uncertainty there; surely someone out there has done this properly!

So, when is the PDO considered to have changed phase, according to the professionals?

When is the next change, from negative to positive; when is this likely to occur?

And what will be the effect of this on the global average temperature trend?