Divergence and reconvergence of UAH and HADcru

Posted: July 11, 2010 by tallbloke in solar system dynamics

On WUWT: Christopher Hanley says:
July 10, 2010 at 11:31 pm (Edit)

HADCRUT3 and UAH noticeably diverge between 1980 and 1997 which is mysteriously corrected post 1997:

Divergence and reconvergence of surface and tropospheric trends.

Thanks Chris for posing a good question and making a useful graph.

The first thing to note is that the divergence amounts to around 0.125C. If the oceans were to release enough energy to become 0.125C colder all at once, the tropospheric temperature would increase to over 100C and we would all be instantly prawned. This is not a trivial amount of energy.

I think there is at least partly, perhaps mainly a physical reason for this divergence/reconvergence in the datasets. It’s mainly to do with cloud cover. ISCCP shows diminishing cloud cover from 1979-1998.The Earthshine Project (Palle et al) shows increasing cloud cover from 1998.

During the less cloudy years, the ocean and land absorbed higher levels of surface insolation and the surface temperature datasets showed an increase at a higher rate than the tropospheric temp, mainly because the ocean sequestered a lot of the heat in it’s absorption phase and therefore had an increased surface temperature. The troposphere lost heat more readily through less cloud cover, hence the lower UAH trend.

Then after 1998 it got cloudier, the land and ocean absorbed less heat so surface measurements slowed their rate of increase. But the ocean has started releasing the sequestered heat now cloud has increased and the less active sun isn’t forcing more energy in and downwards. This is in part why we got the big el nino in 1998 and another one this last year. The ocean is releasing heat it has held for a long time, as evidenced by the decline since 2003 in ocean heat content which had been rising since the ’60’s. Increased cloud also means more warmth retained in the troposphere during the night-time.

But although increased low cloud cover should increase the amount of downwelling longwave radiation (greenhouse effect), it doesn’t penetrate the ocean beyond it’s own wavelength, so it doesn’t warm it. It just increases evaporation at the surface.

This is why the tropospheric trend has outstripped the surface trend in recent years.
I see this as further backing for my simple Sun-Ocean energy model.

In closing, a look at the UAH and HADsst2GL trends:

You can’t push a pin between the overall 1980-2010 trends for UAH and HADsst2GL. Which I think is probably a testimony to the quality of work from both Hadley and UAH. Well done guys and girls! But then, wait a minute. If the ocean had heated up 0.4C on its surface and was radiating more, wouldn’t we expect the troposphere to rise in temp by more than that, given its lower heat capacity? Expert opinions please.

Discuss! 🙂

  1. You are right. Cloud cover began reducing in about 1991, as you can see at H.Svenamark´s “The Chilling Stars”pp.77, caused by the increased Sun´s radiation, after the big CME events in 1989, one of which produced the Quebec´s grid blackout.
    But…some took care to “adjust” satellites measurements which showed a sudden increase in TSI, as Nicola Scafetta shows it:
    So…as long after Climate Gate showed it, they just made a simple “Trick”.
    As you say: During the less cloudy years, the ocean and land absorbed higher levels of surface insolation and the surface temperature datasets showed an increase at a higher rate than the tropospheric temp, mainly because the ocean sequestered a lot of the heat in it’s absorption phase and therefore had an increased surface temperature

  2. Tenuc says:

    I think your right about the cloud cover changes and the ocean heat content, with the ocean acting as a giant radiator to even out the spiky energy our planet receives in all its different forms.

    The ocean heating/cooling cycle is very complicated and rates of cooling depend on many factors (e.g. cloud cover, amount of UV, wind speed, enthalpy of evaporation, salinity, speed of currents, amount of sea ice e.t.c). I don’t think there is one simple mechanism responsible for the changes we observe, rather it is a combination of many non-linear events and is strongly influenced by the various solar/lunar/planetary quasi-cycles.

    Intuition tells me that if the multitude of processes which comprise our climate is the orchestra then the moon and planets are the lead violinist, with the role of conductor being taken by our sun.

  3. Tenuc: See ¨Holistic Clinate Theory” (left column), and a evidently unbiased paper (as it did not originated in some “developed” country):
    Solar Forcing of the Stream Flow of a Continental Scale South American River

  4. tallbloke says:

    Tenuc, there are many things affecting ocean energy absorption and emission. I’m struggling to get to an overview which cuts through the complexity to enable us to quantify it in a meaningful way. This is needed so we can get a better handle on the length of time the ocean will keep us warm if the sun gets really quiet.

    My work on calculating the ocean energy acquisition from the sun in the 90’s, along with the rate OHC is dropping since 2003 tells me we have enough time to get some organisation into agricultural practice. So long as the policy makers wake up in time.

  5. Tallbloke: As seen in H.Svensmark´s graph (op.cit.pp.77), drop of cloud cover happened about 1991 and the big El Niño in 1997, a six years lag.

  6. tallbloke says:

    Both Pinatubo in the phillipines and and Hudson in the Chilean Cerro erupted in 1991, VE6 and VE5 volcanic eruptions. A double whammy to the troposphere in the equatorial and southern ocean regions. Did the volcanic aerosols assist in the cloud reduction alog with strong solar cycle 22? How? Ozone destruction?

  7. The effect of the areosols, mainly SO2 (as H2SO3 when reacted with water and H2SO4 when reacting with water and oxygen)would be the contrary, as these are seeds for drop-cloud formation. Really these volcanoes diminished the effects of sun´s strong cycle.
    About ozone: One of the possibilities is that when the earth receives UV radiation it produces O3 (ozone) and when under a proton flare (hydrogen nucleii) chances are it reduces ozone changing it to H2O (water), so the water cycle could receive some “help” from above.
    The ancient Incas used to watch the pleiades, high in the andes-where there is no humidity- to see if these were seen “blurry”, if so, it meant that the coming season was going to be a good season due to rain.

  8. tallbloke says:

    It’s true the sea level rise graphs from satellite don’t start until late ’92, so they may have missed the effect of Pinatubo on sea level rise.

  9. Bob Tisdale says:

    tallbloke: How do you account for the shift in the HADSST2 data…

    …that is caused by their merging of two (incompatible) source SST datasets in 1998, which is the likely cause of the change after 1997? Refer to:

    HADSST2 is also spatially incomplete, especially in the mid-to-high latitudes of the Southern Hemisphere. Refer to:

    I would suggest you compare HADISST or Reynolds OI.v2 SST data to UAH TLT anomalies. Are the results different?

  10. tallbloke says:

    Thanks Bob, great to get some input from you at this point. I’ll go and see what KNMI has got.

  11. Bob Tisdale says:

    Beat ya:
    UAH TLT Ocean vs HADISST:

    UAH TLT Ocean Minus HADISST:

    The difference appears to represent the exaggerated response of the TLT anomalies.

  12. tallbloke says:

    Eyeballing your comparison graph http://i46.tinypic.com/2di3iut.jpg it seems that is maybe where the answer to my question about how the trends for SST and lower troposphere could be the same lies. The answer seems to be that Hadley sk(r)ewed up the data splice at 98. Sneaky…

    I always wondered how the SST managed to stay high after the 98 el nino despite a load of energy obviously having been burped out of the ocean into the atmosphere.

    http://i29.tinypic.com/1qnxut.jpg seems to show the best effort difference in the 1979-2010 trend for ocean and lower troposphere is 0.05C then. This is about 0.016C/decade. Which is not a lot, but non-zero.

    Thanks very much for your timely comment.

  13. Tenuc says:

    Thanks, Adolfo. for the link to the paper which has an interesting conclusion:-

    Solar Forcing of the Stream Flow of a Continental Scale South American River
    “Conclusion — Stream flow variability of the Parana river has three temporal components: on the secular scale, it is probably part of the global climatic change, which at least in this region of the world is related with more humid conditions; on the multidecadal time scale, we found a strong correlation with solar activity, as expressed by the sunspot number, and therefore probably with solar irradiance, with higher activity coincident with larger discharges; on the yearly time scale, the dominant correlation is with El Nino.”

    Perhaps comparing solar activity to GMT smears out the signal, which perhaps has a big effects on a regional scale and these effects could differ across the globe. The Parana river discharges on average 640,000 cu ft/s or fresh water into the South Atlantic and peak flow at times of flooding must be much more than this. I wonder how much this changes the local salinity in the Atlantic and what effect this would have on ocean currents?

    Perhaps lots of small events across the globe add up to make significant changes to important climate mechanisms, which over time, effect the whole?