Is this the 4.5 year cycle in ARGO and tropical Atlantic?

Posted: October 9, 2012 by tchannon in Analysis, climate, Cycles, Ocean dynamics, Surfacestation, weather


This is a very provisional quick look at some of the CLIMAT data from UK sites, chosen as probably fairly trustworthy over this time period in the light of the ongoing surface stations project.

I pulled out Tmax and Tmin, fed into the novel analysis software here to see what it makes of it. I’ve let it extrapolate.

The most obvious commonality is the above. It isn’t,large enough to notice to a human, half degree either way.

As expected Tmin is more problematic. I won’t go into the gory details but most of it makes sense in relation to site instabilities.

I expected Lerwick to be a reference so I was dismayed to note it only starts data in 2000. Very little is known about the site.

The inclusion of Lyneham is a hunch, slightly know the place and I suspect it has been fairly consistent without the “little war” panic changes near the met site. (aerodrome is closing end of this year so that dooms the meteorological site)

Manston is problematic but Tmax far less so than Tmin.

Most of remaining stations are active military bases or civil airports. If I do the other stations, needs automating, I have no idea if the above will hold up.

Quick web search, see if this period is known.

To my surprise there is recent work pointing at this period as a precursor to UK weather flows.

“Variability in the Global Equatorial Ocean (1000m) from Argo

Argo data show that 4.5-year Atlantic cycle associated with slowly propagating high-baroclinic mode waves is the dominant variability at depth”
— “Tropical Atlantic Climate Experiment Peter Brandt, IFM-GEOMAR, Kiel, Germany, 10M PDF from


— PDF of abstract above EGU2011-4599 – CO Meeting Organizer

I’d be interested to see the phase of what they found.

The origin of 4.5y, I’d expect this to be related to earth plate positions, geometry of ocean and land with the excitation the annual orbit.

That will do for the moment, more if you search.

post by co-moderator

  1. Coldish says:

    There must have been a climate station at Lerwick before 2000. See

  2. tchannon says:

    Here we are, won’t be the same processing but good enough.
    (be lower noise because sig process methods are used)

  3. suricat says:


    Any chance this cycle was induced by insolation in the blue to Uvb spectra?

    If so, there’s the possibility that it’s a forcing from solar influence (sunspot propensity).

    Best regards, Ray.

  4. suricat says:


    In my last post, for clarity, I was referring to the paper and not the graph. 🙂

    Best regards, Ray.

  5. tchannon says:

    So you want a pretty blue background. Well, there was plenty of swearing.

    I don’t know.

  6. tallbloke says:

    At this frequency I’d expect some interaction with lunar tidal modes. Is there any second order cyclicity of Tmax to be found in the longer record?

  7. tchannon says:

    Nothing else is common looked at like that. I seem to be monosyllabic at the moment.

  8. Paul Vaughan says:

    1) What is your most precise estimate of the period? (Is it 4.5000000 years?)
    2) What are the lengths of the records from which you derive your estimates?

  9. tchannon says:

    Excluding Lerwick.

    This will look confusing
    Period_years | phase_radians (relative to start date) | amplitude (1/(2.sqrt(2)) for p-p)
    4.62 4.99 0.44 Eskdalemuir
    4.46 4.71 0.5 Lyneham
    4.51 4.6 0.49 Manston
    4.38 4.35 0.33 Camborne

    One northern high wet inland, three across the south, Atlantic to SE coast.
    Jan 1996 to Sept 2012 by month

  10. Paul Vaughan says:

    “I’d be interested to see the phase of what they found.”

    Maybe check directly
    via visual inspection of the graphs included here
    or directly from Argo data.
    Also note that they give geographic coordinates for surface expression (in equatorial Atlantic wind & SST). Maybe use KNMI Climate Explorer to see if other series record the cycle (and if so measure the phase).
    Please update us with any new period estimates (needed to refine speculation about drivers) and with any further discoveries of non-equatorial-Atlantic expression.