Lamenting the Double ITCZ Bias in CMIP5 Climate Models

From CO2 Science, something that Willis Eschebach holds dear as a regulating mechanism: The Inter-Tropical Convergence Zone

Circulation in the ITCZ

Circulation in the ITCZ

In what must be a difficult acknowledgement to make, Oueslati and Bellon (2015) write that “the double intertropical convergence zone (ITCZ) bias still affects all the models that participate in CMIP5,” – i.e. the Coupled Model Intercomparison Project, phase 5 — while further noting that “as an ensemble, general circulation models have improved little between CMIP3 and CMIP5 as far as the double ITCZ is concerned.” And, therefore, they proceed to discuss some of the many other problems that have ultimately led to this problem.

The two French researchers begin by noting that (1) “the double ITCZ bias affecting the central Pacific can be connected to the simulation of a too-zonally elongated South Pacific convergence zone,” as well as (2) “a too-zonally elongated SPCZ and a spurious ITCZ in the Eastern Pacific, that (3) the spatial distribution of sea surface temperature is “poorly simulated in coupled ocean-atmosphere GCMs (OAGCMs),” with (4) “a positive SST bias over the southeastern Pacific,” as well as (5) “an excessive equatorial cold tongue extending too far west in the Pacific,” which latter biases are attributed to coupled ocean-atmosphere feeedbacks such as (i) “the SST-wind-induced surface fluxes feedback,” (ii) “the SST-stratus feedback,” and (iii) “the SST gradient-trade wind feedback associated with vertical upwelling.”

After further studying this sad situation, Oueslati and Bellon additionally, and quite rationally, come to the conclusion that “overestimated ascending regimes suggest that processes inhibiting deep convection (e.g. convective entrainment, downdrafts and large-scale subsidence) are still poorly represented in CMIP5 models,” all of which makes one wonder if it will ever be possible to correctly represent these several interacting phenomena in a fail-safe climate model.

Paper Reviewed

Oueslati, B. and Bellon, G. 2015. The double ITCZ bias in CMIP5 models: interaction between SST, large-scale circulation and precipitation. Climate Dynamics 44: 585-607. http://link.springer.com/article/10.1007%2Fs00382-015-2468-6

Abstract

The double intertropical convergence zone (ITCZ) bias still affects all the models that participate to CMIP5 (Coupled Model Intercomparison Project, phase 5). As an ensemble, general circulation models have improved little between CMIP3 and CMIP5 as far as the double ITCZ is concerned. The present study proposes a new process-oriented metrics that provides a robust statistical relationship between atmospheric processes and the double ITCZ bias, additionally to the existing relationship between the sea surface temperature (SST) and the double ITCZ bias. The SST contribution is examined using the THR-MLT index (Bellucci et al. in J Clim 5:1127–1145, 2010), which combines biases on the representation of local SSTs and the SST threshold leading to the onset of ascent in the double ITCZ region. As a metrics of a model’s bias in simulating the interaction between circulation and precipitation, we propose to use the Combined Precipitation Circulation Error (CPCE). It is computed as the quadratic error on the contribution of each vertical regime to the total precipitation over the tropical oceans. CPCE is a global measure of the circulation-precipitation coupling that characterizes the model physical parameterizations rather than the regional characteristics of the eastern Pacific. A linear regression analysis shows that most of the double ITCZ spread among CMIP5 coupled ocean–atmosphere models is attributed to SST biases, and that the precipitation large-scale dynamics relationship explains a significant fraction of the bias in these models, as well as in the atmosphere-only models.

Full paper (draft) http://www.cnrm-game.fr/IMG/pdf/ditcz_cmip5.pdf

 

Advertisements

44 thoughts on “Lamenting the Double ITCZ Bias in CMIP5 Climate Models

  1. Maybe it is time to go back to observe what is really going on in nature rather than rely on models. We must fully understand the importance of clouds and the role they play in the daily, seasonally, and long term climate stabilization, putting a cap on maximum temperatures.

    • Totally agree but even then the “scientists” go in and adjust the facts to fit their narratives, THAT is the problem with the current BS that now even the Pope is spouting. Disgusting.

      • Require scientists to follow in the footsteps of Kepler, Newton, and Einstein. Require them to develop laws, well confirmed hypotheses, that specify the natural process in question. Use Kepler’s Three Laws, which are purely empirical, as the example. Then the observable facts are specified and cannot be endlessly adjusted.

      • While I like the idea, the weakness – and it is a big one – is that empirical data is generally capable of supporting several different explanatory attempts with roughly equal efficiency. Consequently, when your “empirical” data contradicts an explanation (law) that operates very efficiently there is a general tendency to: first, question the data, second, patch the explanation, and only last return as far towards the basics as necessary to develop a “clean” explanation that works as well as the original but admits the results of new empirical data. This is simply how science and scientists work. One doesn’t reinvent the wheel unless one has determined that the existing one really needs a redesign – insufficiently round, lack of hub location for axle, etc.

    • The Meteorology and Climate of Tropical Africa, Marcel Leroux… http://www.praxis-publishing.co.uk/9783540426363.htm
      He described very well the complex geometry of the Meteorological Equator as opposed to the simplistic illustrations found everywhere, see figure 4 of this publication.
      http://ddata.over-blog.com/xxxyyy/2/32/25/79/Leroux-La-dynamique-de-la-grande-secheresse-sahelienne.pdf
      Leroux addressed those inhibiting features of the ITCZ. Perhaps when models will take notice of observations… “overestimated ascending regimes suggest that processes inhibiting deep convection (e.g. convective entrainment, downdrafts and large-scale subsidence) are still poorly represented in CMIP5 models,”

      • Tom Rude:
        I welcome your reference to Leroux’s magnum opus, which is the product of decades of truly scientific field experience in the region, not just armchair theorizing or academic modeling. Half a century earlier, very similar depictions of mechanisms were provided in Hamilton and Archbold’s long paper on the meteorology of Nigeria. Only at WUWT is the role of deep tropical convection in regulating surface temperature touted as Willis’ thunderstorm hypothesis.

      • Woaw checking amazon Canada, the 2010 book is at $86 CAD… really a deal.
        Dynamic-Analysis-Weather-Climate-Perturbations

    • Right. Require climate scientists to focus on the several natural processes that make up some phenomenon such as cloud behavior.

    • Oh, come on, don’t be silly!
      This is climate science we’re discussing!
      “The data doesn’t matter. We’re not basing our recommendations on the data. We’re basing them on the climate models.”
      ~ Prof. Chris Folland ~ (Hadley Centre for Climate Prediction and Research)

      • So what do you do when the models do not agree with the measurements? According to the models there should be an equatorial hotspot in the troposphere, but it has been found elusive. Do you ignore the data, or do you acknowledge the models to be incomplete or misleading? IPCC has chosen to ignore the data, yielding to political pressure. Is that good science?

      • ““The data doesn’t matter. We’re not basing our recommendations on the data. We’re basing them on the climate models.”
        Steve Mosher agrees, “you cant understand without models.
        understanding itself IS A MODEL”, …thus Steve M defends ignoring data that contradicts your theory.

      • I agree. But running a bad model ten thousand times does not improve our understanding of the climate in any way. Dame Slingo has a nice picture regarding an irreducible uncertainty of models – but apparently you can run any model, however imprecise, 10,000 times, to get an understanding .. of what, Steven? Of a behavior of a model, or of climate?

      • Steven Mosher
        June 17, 2015 at 10:11 pm
        you cant understand without models.
        understanding itself IS A MODEL

        ——————————————–
        Quite true, but, and this is very serious, with a model that “works,” you can slide into a persistent state of self deception and close your eyes to the warts on the hog. No existing model of “climate” works sufficiently well enough to remain within shouting distance of empirical reality. The maintainers tend to accept the CO2 “governor” assumption and cannot make the current models approximate reality without reducing the influence of CO2 to where it is an unimportant participant in “climate.” Current Climate Science has descended into a maze of Ptolemaic thinking, attempting to patch models until they “work.” It has done so for the very same reasons that Ptolemaic astronomy did so. The presumption that the basic assumptions about how the universe works were correct. They were not.
        When you look at climate, the “basics” is weather, and climate is “understood from” and studied through weather. Climate does not “drive” or “cause” weather. Quite the contrary, “scientists” generalize climate from weather. Climate is nothing but an historical summation of weather at any time span. We can talk about the climate during the ice age, but what changed was the weather. Since the geological data resolution is so poor, we literally see a blurred image of “climate” change instead of ice age weather. That blurred view becomes reified as we try to discuss Ice Age “climate.” The core error in climate science is that “climate” exists beyond weather in some fashion.

      • Steven Mosher: “you cant understand without models.
        understanding itself IS A MODEL”

        Inded.
        The trick is to know which models work and which don’t.
        You need to work on that.

      • lenbilen: “So what do you do when the models do not agree with the measurements?”
        You adjust the measurements until they match the data of course!
        What else?

      • Oops!
        That should be “You adjust the measurements until they match the MODELS of course!

  2. There’s nothing wrong with using models. We use models all the time. Finite element and fluid dynamics models have driven much innovation. Circuits are modeled to give us all the chips that drive all our gadgets. There is a model of the physical response of your car running in your stability and control system (any closed loop control system reflects a model of the real system).
    The problem is using models that haven’t been validated, haven’t been shown to have predictive power, and show important differences with the observed system for decision making purposes. Using the model results where sub-scale (i.e. local) results aren’t even close to predictive to correlate the global results to local events in the real system (i.e. weather attribution) is simply indefensible.

    • Yes. Even if the models worked well, trying to work on the complexity of CFD for climate at a planetary scale is going to take very much better computers than anything we are going to have in the foreseeable future plus collation of a very much greater data set to fix a reasonable start point.

    • Well, yeah, give up the illusion that one must model the entire climate before one can model a part of it.

    • Those models are actually based on science. The derivation of each of the models is documented, with all known assumptions are documented. The computer codes written to implement the models are well documented, including any additional assumptions made to get the code to work on existing computers. Mathematical techniques are identified. The models are validated as being representative of the real world, and the codes are verified as being correct implementations of the model. That’s why engineers regularly use both FEMs and CFD codes, and trust the results – they know the limitations of the models and the codes.

      • Mosh: Yes engineers do that and they validate them using predictions and measured confirmations of them. That’s why bridges don’t fail and forecasts of climate disasters do.
        The climate model forecasts made using a high sensitivity to CO2 have been invalidated by subsequent measurements. Real engineers admit that.

    • Actuality we will never have the computer power to model climate, the variables are far to many and to do it on a scale fine enough to get a accurate ideal of what might be going on will take and computer now or in the foreseeable future longer than the universe has been around. Modeling climate in the fashion we model other things is never going to work, it a fools errand and should be treated as such.

    • “The problem is using models that haven’t been validated”
      I think it is worse then that. They are using models that have been invalidated by observations.

  3. Here is a paper from 2 years ago:
    “Based on our energetic analysis, cloud biases over the Southern Ocean are the main cause of biases in atmospheric cross-equatorial energy transport and the tropical precipitation asymmetry index. In addition to their effect on cross-equatorial energy transport and tropical precipitation, these cloud biases have also been shown to lead to biases in jet stream latitude (27) and total meridional energy transport (28).”
    Link between the double-Intertropical Convergence Zone problem and cloud biases over the Southern Ocean
    http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3612620/
    Published online 2013 Mar 14

  4. The question then is, when they speak of a ‘positive SST bias’ are they talking about the instruments or the adjustments?
    Would that we had a world where this question didn’t have to be seriously asked about people that take readings from instruments.

  5. You mean ‘climate science’ is only just figuring out the problem??!
    As Ethically Civil points out (and my experience in another branch of engineering mirrors EC’s) all this means is that the CAGWarmingWarriors are just getting the idea that they need to validate their models.
    All they have done so far is to ‘invalidate’ the models or, in the case of Karl et al 2015, to fineagle the data to decrease the quantum of invalidation, which is hardly a robust, scientific way of going about these things.

    • “You mean ‘climate science’ is only just figuring out the problem??!”
      What problem? The science is settled.
      Right? Right???

  6. As far as I can recall Willis’s hypothesis is limited to convective activity in the tropics.
    In contrast I have, since 2008, been pointing out that latitudinal climate zone (and ITCZ) drifting together with changes in jet stream zonality / meridionality is the main negative system response to any forcing element.
    Referring to the northern hemisphere:
    “If jet streams, on average, are further south then the high pressure systems to the north of them predominate and the globe is cooling. If, on average, they are further north then high pressure to the south of them predominates and the globe is warming.”
    from here:
    http://www.newclimatemodel.com/weather-is-the-key-after-all/
    June 18, 2008

    • Stephen, if you look at the actual energy budget, the Temperate and Polar zones don’t really matter except for radiating out the energy picked up in the Tropics. The jet streams energy impact isn’t measurable and CO2’s impact even less so.
      Over 84% of the earths energy budget is in the Tropics, that is why Willis focuses on the Tropics. It is best to pick the low hanging fruit first.

      • jinghis,
        I find that the poles are equally important because the top down solar effect on tropopause height via changes in the ozone creation.destruction process is greatest at the poles.
        Climate change is simply the ever changing interaction between the top down solar effect at the poles and the bottom up ocean effect at the equator.

      • Stephen,
        There is very little ‘top’ down solar effect at the poles. Solar panels are placed upside down in fact, because the sun never rises very high and the reflection is higher. The Sun basically circles the horizon.

  7. The climate models also do not model well the “left over” and “moving” warm and cool ocean pools and strips of water at various strata on a global scale. These oceanic pools and strips of water migrate, rotate, and overturn. What was incoming solar heat in one place is now showing up as deeper heat or shallower heat in another place on the globe. What was a pool is now a strip of water, etc. The paper seems to focus only on the atmospheric model weaknesses. That’s like trying to walk on one leg without at least a crutch for the missing leg.
    Modeling the interconnected intrinsic drivers of Earth’s very powerful and natural oceanic and atmospheric weather pattern variations over short and long term time spans is a wickedly complex problem. Given that realization as fact, it amazes me that anthropogenic warming researchers would think it entirely appropriate to model anthropogenic greenhouse warming by adding a fudge factor to the mix. It reminds me of young children who think they can fly simply by adding a cape to their outfit.

    • Very well said. Though I would call it “rapture of the model.” The larger the model the more the rapture.

    • Don’t worry, the Nobel Commitee will be contacting the Hockey Team about picking up their Prize for solving the Navier-Stokes equations and the turbulent flow problem, after that the Hockey Team will begin work on the cold fusion.

  8. Well, under these circumstances the standard climatological procedure is to adjust observations until they show an undeniable split intertropical convergence zone. They also need to remove from all measured datasets the unbelievably high level of interhemispheric symmetry in annual average absorbed shortwave radiation, never replicated by any computational climate model.
    That’s the way forward, is it?

Comments are closed.