Comparative study of the GEOMAR illustrates dependence on model resolution
Climate models can only provide an approximation to the real world. Some models still suffer from relatively large biases in the sea surface temperatures of the tropical Atlantic. A new study by GEOMAR Helmholtz Centre for Ocean Research Kiel shows that in a model with higher resolution the errors in the tropical Atlantic are much smaller relative a model of the same family employing coarse resolution. In addition, climate change projections with the high-resolution model show more pronounced changes in atmospheric circulation and precipitation in this region. The results have now been published in the international journal npj Climate and Atmospheric Science.
A long-standing problem in global climate models (coupled ocean-atmosphere-sea ice models) is the too warm sea surface temperatures in the eastern tropical Atlantic, which also influences the nature of the interactions between the ocean and the atmosphere. Simulations at GEOMAR Helmholtz Centre for Ocean Research Kiel show, on the one hand, that temperature errors are much smaller when using in a climate model an atmospheric component with very high resolution. On the other hand, only such a climate model projects a sea-surface temperature response to increasing atmospheric CO2-concentrations that is similar to the warming trends observed since the mid-20th century. Moreover, the atmospheric circulation pattern at the equator, the so-called Walker circulation, is significantly altered in the climate change experiment, leading to greatly increased precipitation in the equatorial region. The scientists have now published the results of their investigations in the international journal npj Climate and Atmospheric Science.
“We have used the Kiel Climate Model in two configurations”, explains lead author Dr. Wonsun Park from GEOMAR. One uses an atmosphere model with a coarse horizontal resolution of just under three degrees (~300 km) and 31 vertical levels, the other a horizontal resolution of about half a degree (~50 km) and 62 levels”, Park continues.
The higher atmospheric resolution greatly reduces the error in the simulated sea surface temperatures of the tropical Atlantic. “This is due to the much better simulation of the surface winds, which is not possible with the coarse-resolution model because the transitions from land to sea regions cannot be represented well enough”, explains co-author Prof. Dr. Mojib Latif from GEOMAR. “The surface winds have a considerable influence on the ocean currents and thus on water temperatures”, explains the climate scientists from Kiel. Thus, according to the authors, the high-resolution atmospheric component configuration of the climate model possibly may provide enhanced future projections in the tropical Atlantic region relative to most other climate models that typically employ a coarse-resolution atmospheric component.
Both model configurations were used in a climate change experiment in which atmospheric carbon dioxide concentrations continue to increase. The response of the two model versions is very different. The high-resolution variant reacts much more sensitive, the atmospheric circulation pattern in the tropics changed fundamentally, the hydrological cycle is increased. In contrast, the coarse-resolution version shows a much weaker atmospheric reaction. “We were somewhat surprised by the clear difference”, says Professor Latif. “But it clarifies that our understanding of all the developments associated with climate change is still immature”, Latif concludes.
Park, W., and M. Latif, 2020: Resolution dependence of CO2-induced Tropical Atlantic sector climate changes. Climate and Atmospheric Science, doi: https://doi.org/10.1038/s41612-020-00139-6
But it clarifies that our understanding of all the developments associated with climate change is still immature”, Latif concludes.
Yup, the science is settled!
Willis E has shown this in an other way.
Negative feedback from increasing radiation:
Rain and hurricanes will eat up the energy added.
That is the most telling sentence of the whole article, you can not effectively model something unless you have a complete understanding of it. Which the climate alarmists don’t! That is why these models must not be relied upon for policy because far more damage could be done then any alleged benefit. But then the politicians who push this nonsense have an agenda, and it is not the Climate!
“Both model configurations were used in a climate change experiment in which atmospheric carbon dioxide concentrations continue to increase.”
WTF? FFS! Do these bozos not understand that a COMPUTER SIMULATION is NOT AN EXPERIMENT!
“Do these bozos not understand that a COMPUTER SIMULATION is NOT AN EXPERIMENT!”
That was my first reaction to the article also.
I also found it interesting that they considered “a horizontal resolution of about half a degree (~50 km) and 62 levels” a “high-resolution variant”.
It’s a hypothesis masquerading as a theory in the age of plausible (e.g. pattern matching, inferential) science, but their models… our skill to perceive, forward and backward, is still constrained to a limited frame of reference. We really need a separation of logical domains. Science begins (e.g. hypothesis) life in the philosophical logical domain, then after rigorous application of the scientific method, either crosses to the scientific logical domain, remains in the philosophical logical domain, or is shunted to the fantasy (i.e. improbable) or faith (e.g. moral philosophy based on axioms of individual dignity, value, and worth) logical domains.
Can’t agree. I interpret this as an experiment in the reconfiguration of models. Clearly demonstrating, were demonstration needed, that the more you reconfigure your model, the more likely you are to achieve the result that fits your agenda.
“We were somewhat surprised by the clear difference.” This is a signal to revisit all of the “experiment” parameters, not to just accept the surprise results which were in the direction you wanted all along. Science is obviously being ignored here, but that’s where the funding is.
A coarse grid is worse than a finer grid? What a surprise 🙂
Their “fine” grid is 50 km and 62 vertical levels. Does anybody know what grid size is used for – still uncertain, but valuable – hurricane predictions?
Am I surprised? – No!
When the smaller spikes of warm and cold are filtered out due to low sampling dynamics and rate (large bins), the extremes are discarded and you have a more narrow band of temperature range. If there are more small warm spikes than of small cold spikes, the whole picture turn slightly warmer.
The solution: More money for the researchers to enhance resolution, to keep the scare growing.
That would be only a temporary solution. A true solution needs computers a trillion times faster than what’s avaliable now.
The big question is how much of an influence does the tropical Atlantic have if it is 1 deg warmer or colder anyway compared to what comes out of the Arctic ocean when the Beaufort Gyre lets loose and dumps very cold fresher water into the North Atlantic? Throw in a big volcano going off and I wouldn’t want to live in Europe for a couple of decades, especially when all the windmills freeze.
Sooner or later (probably sooner) we are going to have to pay the piper for all these warmer temperatures dissipating and the Sun cycling back to not being so active. The growing seasons are already getting shorter and snows in both hemispheres are making a comeback. These people can model all they want but what really counts is the real world, not the modeled one.
Or as one of the lady’s once, what happens when the battery starts charging again?
“The surface winds have a considerable influence on the ocean currents and thus on water temperatures”,
i.e. non-radiative processes play a significant role at the surface which diminishes radiation’s share and negates BB.
“The high-resolution variant reacts much more sensitive, the atmospheric circulation pattern in the tropics changed fundamentally, the hydrological cycle is increased.”
This means that the adhoc tuning of the water vapour amounts in the tropics will have to be adjusted. Presumably in the coarser model, the water vapour amount was always tuned to match the real world measurements. Since their high resolution model produces more water vapour, they have a problem. Any variable that depends on the amount of water vapour will then be in larger error. Their tuning job is now immense. They will have to tune 100’s of variables just to get the right amount of water vapour. Rather like pressing in 1 side of a balloon which changes the pressure on the other side which then has to be corrected but upon correction changes the pressure somewhere else and so on and so on. Models can never accurately model the climate system because it is a non linear chaotic system as the IPCC has admitted.
With four parameters I can fit an elephant,
with five I can make him wiggle his trunk.
John Von Neumann
It doesn’t matter what resolution you use, the uncertainty in the models make them useless for what they are being used for. If the cumulative uncertainty is 5deg then trying to determine a 0.5deg difference is meaningless.
New and improved simulations
Still completely erroneous
But it clarifies that our understanding of all the developments associated with climate change is still immature”, Latif concludes.
Climate change is the null hypothesis? The models demonstrate no skill to match past observations, let alone foretell the future on regional and global scales. In a chaotic system, where accurate perception is constrained to a limited frame of reference (i.e. scientific logical domain), it’s immature to claim a prognostication, to assume/presume then infill with black… brown matter, where the natural distribution is shifting in time and space and large (e.g. 10, 20, 40, and greater temperature swings in a short period) internal variability.
Somebody wants a new supercomputer.
The high-resolution variant reacts much more sensitive
Resolution is one deficit of the models. The other glarinng deficit is insufficient characterization of the system and processes. So, they improved the precision, but not accuracy of the hypothesis (e.g. model).
I may be foolish but, I still think that 15,000 mile long rift zone running through the middle of the Atlantic Pond needs to be considered further.
…In contrast, the coarse-resolution version shows a much weaker atmospheric reaction. “We were somewhat surprised by the clear difference”, says Professor Latif…
I’m “somewhat surprised” you didn’t expect to see a “clear difference.”
…“But it clarifies that our understanding of all the developments associated with climate change is still immature”, Latif concludes…
What? The sciences is supposed to be settled. Our understanding of ALL the developments associated with climate change is still immature?
Improved models show stronger climate change signal in the tropical Atlantic
how can they claim it’s an improved model? It would have to show improved skill at predicting real world measurements to show its an improved model.. sorry don’t buy it
“Warming oceans more ‘stable’ and that’s bad, scientists warn ”
“Global warming is making the oceans more stable, increasing surface temperatures and reducing the carbon they can absorb, according to research published Monday by climate scientists who warned that the findings have “profound and troubling” implications. ”
One of those “scientists” is our beloved Mickey Mann!
Micky is funny:
Yes Mickey, you finally got, now we just need to get Al to acknowledge that with rising temperature you increase atmospheric CO2, not the other way around – That has really taken some years to sink into your clever heads.
Models- that is about all I needed to stop.
Irrespective of resolution garbage in still results in garbage out.
Characterization and resolution. The value of the latter can only be determined after complete or sufficient discernment of the former. Order matters. Their models are deficient on both points.
GREETINGs FROM NEW Zealand…….did you say Mickey Mouse?? regards, Trevor.
If the model is crap it doesn’t matter how high the resolution is, it is still crap.
Atlantic hurricanes/cyclones today:
Atlantic – Caribbean Sea – Gulf of Mexico
Tropical Weather Outlook (en Español*)
200 PM EDT Mon Sep 28 2020 Tropical Weather Discussion
1805 UTC Mon Sep 28 2020
There are no tropical cyclones in the Atlantic at this time.
And none in the Indian Ocean and very few tropical storms/typhoons in the Pacific Basin presently, except for “severe Tropical Storm Kujira, and is well east of Japan and is tracking to the north but will be turning toward the northeast on Tuesday, local time. This track will steer the storm well clear of Japan and any other land masses before the system loses its tropical characteristics due to increasing shear and cooler waters on Wednesday, local time.” All the recent hurricanes and typhoons have exhausted a lot of tropical heat poleward and off to space. But I doubt this storm season in the NH is over…just having a rest.
Looks like New Zealand is getting hit with climate change as spring sets in in the SH, … https://earth.nullschool.net/#current/wind/surface/level/overlay=temp/orthographic=-194.42,-52.32,826/loc=167.617,-50.949
They are getting a blast of Antarctic air aimed directly at their islands. That may bring them some record cold temps.
Climate models can only be considered “improved” when their accuracy at predicting future states of the Earth’s climate system improve, and not when they produce one or another favored result. That means they cannot possibly improve until a rigorous regime of falsification testing is implemented. Black hat testing, boundary condition testing, the whole works needs to be thrown at these silly toys. Until that happens, climate modeling will remain circular reasoning on steroids, all conclusions tendentious and pointless. Without falsification testing, “improving” climate models is scientifically impossible. And when a Climate Scientologist tells me that “that’s not how we do things” then he’s basically telling me that I should not have any respect for him (or her, hi Cathering Hayhoe, hi Brenda Weurtzal, hi Naomi Oreskes).
Here is the perfect analogy. Climate modeling in the 2020s is merely an exercise in self-pleasuring while sitting in a circle of other self-pleasuring climate modelers, i.e. a giant, very expensive circle jerk.