And finally, AR6 WG1 recognized the failure of AOGCMs to properly handle clouds (which you exemplified above) and noted that about half produce estimates of current warming that are inconsistent with observations. They decided evidence form models was no longer suitable for quantitatively estimating climate sensitivity. Section 7.5.6 and Figure 7.19. (:)) However, everywhere else in the report other authors rely heavily on the output from climate models.
If AOGCMs are ruled out, estimates of climate sensitivity from other lines of evidence might be worthy of examination, sun as Figure 7.12. Of course, all six studies cited rely on models to some extent too.
True Frank. There are quite convincing reasons why climate models can't be even close to predicting patterns of change. Their resolution is way too course, probably at least 50 times too course. As pointed out in this post, their subgrid models appear inadequate as well.
David: At a minimum, you might want to glance at Figure 1.19, the evidence the authors cited before saying AOGCMs can no longer be trusted to provide useful quantitative information about climate sensitivity, which is a pretty revolutionary development. As best I can tell, they probably should be scaling down use of warming from the multi-model mean by an appropriate factor elsewhere in the report.
Thanks Frank, I will try to take a look later this week. I thought I had noticed that the strident and dishonest defense of climate models seems to have died out and my post provoked no climate scientist responses. Perhaps even activists can only deny the obvious for several decades as younger and more honest scientists enter the fray.
Just did a superficial read Frank. Finally the IPCC has done something that is scientifically honest. I applaud them for that.
Comparing CMIP5 and CMIP6 to me shows how difficult it will be to improve climate models. CMIP6 had allegedly improved cloud models. They might be more physically plausible, but they are obviously not interacting with other components correctly.
I learned from a Tapio Schneider talk that all climate models produce far too few marine boundary layer clouds (the most cooling clouds on the planet). He admits this is a well-known problem no one likes to talk about. Sometimes 50% too few. There must be enormous compensating error elsewhere. If the albedo of your model is too low, that problem can be superficially fixed by slightly reducing the saturation percentage needed to make clouds start forming. (This parameter is less than 100% because in the real world an entire grid cell doesn't need to be saturated before some clouds start forming because the contents of a grid cell aren't perfectly homogeneous.) IMO, If your model can't produce realistic clouds where they are observed, then there is no hope of properly representing changes in clouds (cloud feedback). IPCC reports show us how well models reproduce observed temperature, diurnal temperature change, seasonal temperature change, precipitation and a lot of other observables everywhere on the planet, but I've never seen a figure showing well they reproduce clouds.
You'd think cloud-resolving models would help AOGCMs do better modeling MBL clouds. Unfortunately, MBL clouds are not created by a local process: Descending dry air originates on the other side of an ocean basin. The cold ocean below is the result of currents heading from the poles to the equator and from upwelling of cold water from below, driven by lower weight of surface water from the planet's rotation and by prevailing winds. Clouds form in the complicated interface between the boundary layer and the free troposphere. This is a vastly more complicated than clouds forming by simple upward convection.
Caltech has network of very rich donors they rely upon to get high-potential high-risk research programs to the point where they can attract government or business funding. I believe Schneider is working on models with very small grid cells embedded in critical locations.
I'm glad they are working on it but as I said in my blog post there are giant hurdles to overcome here including some challenging theoretical ones. I think that for the foreseeable future, climate models won't tell much of use. Better to focus on better data where we know how to proceed.
David Young may be an expert at aerodynamics but not at real fluid dynamics, such as ocean water with a delineated thermocline. Variable torqueing applied to the Pacific ocean results i ALL natural variability.
You don't know my work Paul. If you read my blog post you will see that solving the Navier-Stokes is the same in aerodynamic simulations or the oceans or the atmosphere. That's according to Nic Stokes too.
Pukite: Debates over who is an expert are irrelevant now that the IPCC itself admits that evidence from climate models is no longer considered reliable enough to be used to estimate climate sensitivity. As linked above, they show that fully half of models seriously overestimate current warming. And the best estimates of climate sensitivity based on currently warming and forcing change are consistent with a climate sensitivity from 1-3 degC, something the IPCC can only explain by postulating global warming over the past four decades has been suppressed by unforced variability.
I had a post at Climate Etc. in December of last year going through the CFD literature and explaining why this lack of skill is implied by the literature.
Parameters are best guess ranges, not exactly robust science, not reality of what's happening. When the models use best guesses, it's not an attribution study, it's simply adjusting the algorithms to what they want to see or expect to see. Then they attribute the "results" to the forcings and feedbacks of their best guess ranges. Maybe that's why the models aren't very successful.
Simply simulation games that reside more in sci-fi than reality. No models projected the 20 year hiatus, we're also into a 9 year cooling trend. The models aren't designed with reality in mind.
And finally, AR6 WG1 recognized the failure of AOGCMs to properly handle clouds (which you exemplified above) and noted that about half produce estimates of current warming that are inconsistent with observations. They decided evidence form models was no longer suitable for quantitatively estimating climate sensitivity. Section 7.5.6 and Figure 7.19. (:)) However, everywhere else in the report other authors rely heavily on the output from climate models.
If AOGCMs are ruled out, estimates of climate sensitivity from other lines of evidence might be worthy of examination, sun as Figure 7.12. Of course, all six studies cited rely on models to some extent too.
True Frank. There are quite convincing reasons why climate models can't be even close to predicting patterns of change. Their resolution is way too course, probably at least 50 times too course. As pointed out in this post, their subgrid models appear inadequate as well.
David: At a minimum, you might want to glance at Figure 1.19, the evidence the authors cited before saying AOGCMs can no longer be trusted to provide useful quantitative information about climate sensitivity, which is a pretty revolutionary development. As best I can tell, they probably should be scaling down use of warming from the multi-model mean by an appropriate factor elsewhere in the report.
Thanks Frank, I will try to take a look later this week. I thought I had noticed that the strident and dishonest defense of climate models seems to have died out and my post provoked no climate scientist responses. Perhaps even activists can only deny the obvious for several decades as younger and more honest scientists enter the fray.
Just did a superficial read Frank. Finally the IPCC has done something that is scientifically honest. I applaud them for that.
Comparing CMIP5 and CMIP6 to me shows how difficult it will be to improve climate models. CMIP6 had allegedly improved cloud models. They might be more physically plausible, but they are obviously not interacting with other components correctly.
I learned from a Tapio Schneider talk that all climate models produce far too few marine boundary layer clouds (the most cooling clouds on the planet). He admits this is a well-known problem no one likes to talk about. Sometimes 50% too few. There must be enormous compensating error elsewhere. If the albedo of your model is too low, that problem can be superficially fixed by slightly reducing the saturation percentage needed to make clouds start forming. (This parameter is less than 100% because in the real world an entire grid cell doesn't need to be saturated before some clouds start forming because the contents of a grid cell aren't perfectly homogeneous.) IMO, If your model can't produce realistic clouds where they are observed, then there is no hope of properly representing changes in clouds (cloud feedback). IPCC reports show us how well models reproduce observed temperature, diurnal temperature change, seasonal temperature change, precipitation and a lot of other observables everywhere on the planet, but I've never seen a figure showing well they reproduce clouds.
You'd think cloud-resolving models would help AOGCMs do better modeling MBL clouds. Unfortunately, MBL clouds are not created by a local process: Descending dry air originates on the other side of an ocean basin. The cold ocean below is the result of currents heading from the poles to the equator and from upwelling of cold water from below, driven by lower weight of surface water from the planet's rotation and by prevailing winds. Clouds form in the complicated interface between the boundary layer and the free troposphere. This is a vastly more complicated than clouds forming by simple upward convection.
Caltech has network of very rich donors they rely upon to get high-potential high-risk research programs to the point where they can attract government or business funding. I believe Schneider is working on models with very small grid cells embedded in critical locations.
https://clima.caltech.edu/our-team/
I'm glad they are working on it but as I said in my blog post there are giant hurdles to overcome here including some challenging theoretical ones. I think that for the foreseeable future, climate models won't tell much of use. Better to focus on better data where we know how to proceed.
David Young may be an expert at aerodynamics but not at real fluid dynamics, such as ocean water with a delineated thermocline. Variable torqueing applied to the Pacific ocean results i ALL natural variability.
You don't know my work Paul. If you read my blog post you will see that solving the Navier-Stokes is the same in aerodynamic simulations or the oceans or the atmosphere. That's according to Nic Stokes too.
Pukite: Debates over who is an expert are irrelevant now that the IPCC itself admits that evidence from climate models is no longer considered reliable enough to be used to estimate climate sensitivity. As linked above, they show that fully half of models seriously overestimate current warming. And the best estimates of climate sensitivity based on currently warming and forcing change are consistent with a climate sensitivity from 1-3 degC, something the IPCC can only explain by postulating global warming over the past four decades has been suppressed by unforced variability.
The link to my blog post is here:
https://judithcurry.com/2022/12/02/colorful-fluid-dynamics-and-overconfidence-in-global-climate-models/
I had a post at Climate Etc. in December of last year going through the CFD literature and explaining why this lack of skill is implied by the literature.
Here's the link.
https://judithcurry.com/2022/12/02/colorful-fluid-dynamics-and-overconfidence-in-global-climate-models/
Parameters are best guess ranges, not exactly robust science, not reality of what's happening. When the models use best guesses, it's not an attribution study, it's simply adjusting the algorithms to what they want to see or expect to see. Then they attribute the "results" to the forcings and feedbacks of their best guess ranges. Maybe that's why the models aren't very successful.
Simply simulation games that reside more in sci-fi than reality. No models projected the 20 year hiatus, we're also into a 9 year cooling trend. The models aren't designed with reality in mind.