Here is a plot from the Australian Bureau of Meterology (BOM) which shows subsurface water temperature in the equatorial Pacific over the last four months.
The sub surface warm pool is melting away. I think this indicates that a Christmas 2012 El Nino event is unlikely. I expect the Aug UAH anomaly will still look fairly warm though, all that oceanic energy has gone somewhere. Into the air is the obvious place. I’ll updat the post when Roy Spencer publishes the Aug UAH graph for the lower troposphere.
H/T phlogiston
Tallbloke's Talkshop 
The daily UAH temperature updates can be found here :
http://discover.itsc.uah.edu/amsutemps/data/amsu_daily_85N85S_ch05.r002.txt
With one day to go, August temperatures are running very close to last years, when the monthly anomaly was +0.33C.
I would be surprised if August 2012 is significantly different.
So that will be 3 La Nina’s in a row. Would 4 in a row be a record, not that Nino / Nina records go back very far, or is there a proxy of some sort that would allow us to track Nina’s /Nino’s back further in time.
It could just stay ENSO neutral for some time.
Cracking ENSO’s code and being able to successfully predict it is a big ambition for many.
Is there any money in cracking the ENSO code?
Yes, and plenty of kudos too. I’m going to persuade Tim C to run the data through his magic software. Watch this space.
TB:
A couple of things to watch out for before Tim C runs the ENSO data through his magic software.
First, ENSO indices are detrended and adjusted for seasonal variations, so a lot of spectral data will be missing or distorted.
Second, all the published SST series except for ICOADS contain a massive stair-step “bucket bias correction” (about 0.5C) in or around 1946 that’s almost certainly bogus.
The closest thing to an uncorrupted ENSO data set would probably be ICOADS after 1950 in the Nino 3.4 area (5N to 5S, 120E to -170E). With a larger area we start getting into time lag issues.
“Second, all the published SST series except for ICOADS contain a massive stair-step “bucket bias correction” (about 0.5C) in or around 1946 that’s almost certainly bogus.”
Actually, you can see the necessity of the correction by looking at other measures.
1. Look at the relationship between land temps and SST ( its a fixed ratio )
2. Look at MAT from the ICOADS database. yes there are just as many MAT measures
as SST measures. look at the relationship between MAt and SST over time by collection
procedure.
If you look at ICOADS you will clearly see the change in reporting for where the measures were taken from ( inlet, unknown etc ) and if you look at the two relations above it will be pretty clear that a correction is needed. The difficulty is figuring out the exact correction.
Hi Mosh. I wonder if there was a biggish el nino in 1945-46 that didn’t get properly recorded due to the reduction of military patrols but a slow restart to merchant shipping in the Pacific. The big ones frequently occur just after solar minimum (1944-5), and often cause a lot of snow and cold weather the year after in northern high latitudes. An old war veteran I spoke to last year was snowed in on his family’s farm in west yorkshire in 1947 – in JUNE !!
I think that would help explain the ‘hump’ and reconcile the datasets.
@Steven Mosher
Click to access final-sst.pdf