climateprediction.net home page
Another cold equator run

Another cold equator run

Message boards : climateprediction.net Science : Another cold equator run
Message board moderation

To post messages, you must log in.

AuthorMessage
Profile geophi
Volunteer moderator

Send message
Joined: 7 Aug 04
Posts: 2167
Credit: 64,403,322
RAC: 5,085
Message 9018 - Posted: 9 Feb 2005, 16:26:01 UTC

Here is a temperature time series from my most recent completed run.
<img src="http://climateapps2.oucs.ox.ac.uk/cpdnboinc/runimage/493191_AT.png">

It definitely had a cold equator and the earth cooled as well. This is the first cold equator run I've had since March, in about 30 completed runs.
ID: 9018 · Report as offensive     Reply Quote
Profile Pete B

Send message
Joined: 26 Aug 04
Posts: 67
Credit: 9,292,129
RAC: 11,653
Message 9068 - Posted: 10 Feb 2005, 9:06:57 UTC - in response to Message 9018.  
Last modified: 10 Feb 2005, 9:12:23 UTC

&gt; Here is a temperature time series from my most recent completed run.
&gt; <img> src="http://climateapps2.oucs.ox.ac.uk/cpdnboinc/runimage/493191_AT.png"&gt;
&gt;
&gt; It definitely had a cold equator and the earth cooled as well. This is the
&gt; first cold equator run I've had since March, in about 30 completed runs.

I have only ever had one true "Cold Pacific Equator" run, that was <a href="http://cpdn.comlab.ox.ac.uk/user/getrun.php?stype=AT&amp;runid=199101">Stacey's</a> last live uploaded run before the offline comparison expt. This was a Classic model initially downloaded/finally uploaded using Amy, hence the credit for the run being to Amy.

It did not redevelop in Phase 3 though, only ever a couple of cells that consistently stayed in the 0-6 deg range.

I did notice though that the 'entrainment coefficient' and 'critical relative humidity' values in my run and your run were the same, although I don't know whether this is significant.

Presumably, these results will decline in number as the team get to know what parameter sets cause them and they are therefore removed from future expts as not representative.

Pete
ID: 9068 · Report as offensive     Reply Quote
Profile old_user17525

Send message
Joined: 13 Sep 04
Posts: 161
Credit: 284,548
RAC: 0
Message 9070 - Posted: 10 Feb 2005, 9:50:08 UTC
Last modified: 10 Feb 2005, 9:50:22 UTC

That last one of mine was a cold equator one and like Geophi's the
whole globe was cold by the time it finished <a href="http://climateapps2.oucs.ox.ac.uk/cpdnboinc/result.php?field=Temperature&amp;resultid=424234&amp;phase=AT#graph">424234</a>

Marj
_________________________________
ID: 9070 · Report as offensive     Reply Quote
crandles
Volunteer moderator

Send message
Joined: 16 Oct 04
Posts: 692
Credit: 277,679
RAC: 0
Message 9072 - Posted: 10 Feb 2005, 10:15:03 UTC - in response to Message 9068.  

&gt;
&gt; It did not redevelop in Phase 3 though, only ever a couple of cells that
&gt; consistently stayed in the 0-6 deg range.
&gt;
Cold equators are rarer in phase 3 than in phase 2 but occasionally you get a cold equator in phase 3 but not in phase 2. (6 of those in the runs used for nature paper).

&gt; I did notice though that the 'entrainment coefficient' and 'critical relative
&gt; humidity' values in my run and your run were the same, although I don't know
&gt; whether this is significant.
&gt;
Those values turn up more frequently in cold equator runs than could be expected just from random chance.


&gt; Presumably, these results will decline in number as the team get to know what
&gt; parameter sets cause them and they are therefore removed from future expts as
&gt; not representative.
&gt;
&gt; Pete

I pursued this idea with DaveF but now I think I was almost certainly wrong.

The Nature paper says these parameters are just having the effect of making the model more variable. More variation in the model causes the cold equators to occur more frequently.

I think it is probably a bad idea to cut out the models that have high variation.

Visit BOINC WIKI for help

And join BOINC Synergy for all the news in one place.
ID: 9072 · Report as offensive     Reply Quote
Profile dubbi

Send message
Joined: 29 Aug 04
Posts: 31
Credit: 155,185
RAC: 0
Message 9102 - Posted: 10 Feb 2005, 21:50:38 UTC

whats This ?? look here


http://climateapps2.oucs.ox.ac.uk/cpdnboinc/runimage/352494_2_temp.png
ID: 9102 · Report as offensive     Reply Quote
old_user23880
Volunteer tester

Send message
Joined: 10 Oct 04
Posts: 223
Credit: 4,664
RAC: 0
Message 9111 - Posted: 11 Feb 2005, 1:58:20 UTC

Your model is showing very strong global cooling in phase 2, which probably means that your parameters are producing an unstable climate. Temps should remain approx level in phase 2. I wouldn't be surprised if your model aborted before the end.

Regarding the culling of unrepresentative models, mentioned earlier by PeteB, my impression from the Nature article was that they are still at the stage of identifying them after completion. Presumably they want to identify bad parameter sets (or part-sets) to remove them from the experiment. Does anyone know whether anyone, apart from Crandles of course, is working on this?
__________________________________________________

ID: 9111 · Report as offensive     Reply Quote
Profile geophi
Volunteer moderator

Send message
Joined: 7 Aug 04
Posts: 2167
Credit: 64,403,322
RAC: 5,085
Message 9114 - Posted: 11 Feb 2005, 2:29:48 UTC - in response to Message 9111.  

&gt; Your model is showing very strong global cooling in phase 2, which probably
&gt; means that your parameters are producing an unstable climate. Temps should
&gt; remain approx level in phase 2. I wouldn't be surprised if your model aborted
&gt; before the end.
&gt;
Mo,

The <a href="http://www.nature.com/nature/journal/v433/n7024/suppinfo/nature03301.html">supplemental material</a> associated with the Nature article gives reasons for the unrealistic cooling seen in some simulations. These may not be "bad" parameter sets, just that the simplified ocean, and a too short calibration phase, with some parameter sets results in the cold equator/cold earth. Running the same parameter sets with the more realistic dynamic ocean of experiment 2 would likely not result in the same unrealistic cooling, so those parameter sets may still be used in experiment 2, even if the results in this 1st experiment cannot be used for much.

George
ID: 9114 · Report as offensive     Reply Quote
Les Bayliss
Volunteer moderator

Send message
Joined: 5 Sep 04
Posts: 7629
Credit: 24,240,330
RAC: 0
Message 9118 - Posted: 11 Feb 2005, 3:39:50 UTC
Last modified: 11 Feb 2005, 3:40:19 UTC

Never mind the cooling. I had that in one model, and a slightly less severe example in another.
What about the EXTREME heating between 1829 and 1830? You can only JUST see the markers
at the top of the screen!

Les
ID: 9118 · Report as offensive     Reply Quote
crandles
Volunteer moderator

Send message
Joined: 16 Oct 04
Posts: 692
Credit: 277,679
RAC: 0
Message 9121 - Posted: 11 Feb 2005, 10:47:29 UTC

I cannot believe the temps in the model really went that high then recovered. I think those temps must just be misreported.
Visit BOINC WIKI for help

And join BOINC Synergy for all the news in one place.
ID: 9121 · Report as offensive     Reply Quote
Profile old_user993

Send message
Joined: 23 Aug 04
Posts: 49
Credit: 183,611
RAC: 0
Message 9322 - Posted: 14 Feb 2005, 22:17:56 UTC - in response to Message 9121.  

&gt; I cannot believe the temps in the model really went that high then recovered.
&gt; I think those temps must just be misreported.

I thought we had one (ages ago - maybe in the beta test) where the thing oscillated - high then low, with larger amplitudes, till it crashed..? So maybe we have a precedent.

Dave
ID: 9322 · Report as offensive     Reply Quote
crandles
Volunteer moderator

Send message
Joined: 16 Oct 04
Posts: 692
Credit: 277,679
RAC: 0
Message 9325 - Posted: 14 Feb 2005, 22:42:53 UTC - in response to Message 9322.  

&gt; &gt; I cannot believe the temps in the model really went that high then
&gt; recovered.
&gt; &gt; I think those temps must just be misreported.
&gt;
&gt; I thought we had one (ages ago - maybe in the beta test) where the thing
&gt; oscillated - high then low, with larger amplitudes, till it crashed..? So
&gt; maybe we have a precedent.
&gt;
&gt; Dave
&gt;

Dave we have had quite a few of these over 25 degree average temperature displays:

<img src="http://climateapps2.oucs.ox.ac.uk/cpdnboinc/runimage/352494_2_temp.png">


which I think is rather different from the oscilation you are talking about. I was wondering if they could be caused by the file size problem I reported in the sulphur cycle alpha test under Windows thread id=50. (That run has crashed so I won't be able to see if this happens.)



Visit BOINC WIKI for help

And join BOINC Synergy for all the news in one place.
ID: 9325 · Report as offensive     Reply Quote
Les Bayliss
Volunteer moderator

Send message
Joined: 5 Sep 04
Posts: 7629
Credit: 24,240,330
RAC: 0
Message 9326 - Posted: 14 Feb 2005, 23:26:59 UTC

After I posted, I looked at the precipitation for phase 2, which showed the same jump.
And there is a spike in both graphs, (phase 2), for the completed graphs, with a severe drop for phase 2.
But both phase 3 graphs are flat.

Perhaps this is one where the 330Meg files need saving!

<a href="http://climateapps2.oucs.ox.ac.uk/cpdnboinc/result.php?resultid=352494"> Result page</a>
Apologies to TermInet&amp;Friends for 'talking behind your back' as it were.

Les

ID: 9326 · Report as offensive     Reply Quote
crandles
Volunteer moderator

Send message
Joined: 16 Oct 04
Posts: 692
Credit: 277,679
RAC: 0
Message 9345 - Posted: 15 Feb 2005, 10:58:47 UTC

TermInet&amp;Friends,

Were there any file size anomalies for that run? (eg the p?29 files being smaller than the same files for different years.)
Visit BOINC WIKI for help

And join BOINC Synergy for all the news in one place.
ID: 9345 · Report as offensive     Reply Quote
Profile dubbi

Send message
Joined: 29 Aug 04
Posts: 31
Credit: 155,185
RAC: 0
Message 9482 - Posted: 17 Feb 2005, 20:18:56 UTC - in response to Message 9345.  

no i dont know but the run has finished and i got some extra points
ID: 9482 · Report as offensive     Reply Quote
Arnaud

Send message
Joined: 3 Sep 04
Posts: 268
Credit: 256,045
RAC: 0
Message 9567 - Posted: 19 Feb 2005, 16:09:25 UTC

Hi,
This is a "little" cold run. In fact I noticed a cold spot in the Pacific Ocean near South America on the vizualisation.
This model has been computed by two users and you can see the little differences on the temp graphs.
My model: <a href="http://climateapps2.oucs.ox.ac.uk/cpdnboinc/result.php?field=Temperature&amp;resultid=429527&amp;phase=AT#graph">click</a>
The other user model: <a href="http://climateapps2.oucs.ox.ac.uk/cpdnboinc/result.php?field=Temperature&amp;resultid=429526&amp;phase=AT#graph">Click</a>
ID: 9567 · Report as offensive     Reply Quote
crandles
Volunteer moderator

Send message
Joined: 16 Oct 04
Posts: 692
Credit: 277,679
RAC: 0
Message 9571 - Posted: 19 Feb 2005, 17:32:15 UTC

<a href="http://climateapps2.oucs.ox.ac.uk/cpdnboinc/result.php?field=Temperature&amp;resultid=490726&amp;phase=AT#graph">Here</a> is my last finished model. There cold equator effect is fairly mild but it is visible on CPView.

An oddly smooth undulating phase 3 graphs.

In parameter terms it is also odd. It has none of the main three parameter setting and one of those is badly set. Consequently I am inclined to think that if this run can generate an effect in both phases 2 and 3, at least half the models could generate these cold equators given sufficiently long simulations.

Of course I am not the first to reach that conclusion. The nature paper got there first saying:

'Some model versions may be more susceptible but any model versions with sufficient variability, potentially including the unperturbed model, could produce it given sufficiently long simulations.'
Visit BOINC WIKI for help

And join BOINC Synergy for all the news in one place.
ID: 9571 · Report as offensive     Reply Quote
Les Bayliss
Volunteer moderator

Send message
Joined: 5 Sep 04
Posts: 7629
Credit: 24,240,330
RAC: 0
Message 9578 - Posted: 19 Feb 2005, 18:30:46 UTC
Last modified: 19 Feb 2005, 18:31:56 UTC

Perhaps models are starting to probe a boundary between stable / unstable.

Chris: your model IS odd. (In a nice sort of way.) Phase 2 is also a bit wobbly. And the effect is very
noticable in the precipitation graph for All.
I hope that not too many of these parameter sets get lost due to people continually crashing models.

Les
ID: 9578 · Report as offensive     Reply Quote
Profile old_user1742

Send message
Joined: 26 Aug 04
Posts: 100
Credit: 1,191,715
RAC: 0
Message 9580 - Posted: 19 Feb 2005, 20:00:48 UTC

Chris, Les, shouldn't we expect about 0.5C variability due to weather. Phase 2 is probably on a slight downward slope with only about 0.2C variation from it. As for phase 3 there are probably perfectly good physical reasons why some models might wave, e.g. increased evaporation -&gt; increased snowfall -&gt; cooler northern summer.

____________________________<br>
<a href="http://www.boincforum.info/boinc/">boinc forum</a> and <a href="http://www.uk4cp.co.uk/">United Kindom</a> team, my climate change <a href="http://www.livejournal.com/users/mike_atkinson/">blog</a>.
ID: 9580 · Report as offensive     Reply Quote
crandles
Volunteer moderator

Send message
Joined: 16 Oct 04
Posts: 692
Credit: 277,679
RAC: 0
Message 9585 - Posted: 19 Feb 2005, 20:31:34 UTC - in response to Message 9580.  
Last modified: 19 Feb 2005, 20:34:06 UTC

&gt; Chris, Les, shouldn't we expect about 0.5C variability due to weather. Phase 2
&gt; is probably on a slight downward slope with only about 0.2C variation from it.
&gt; As for phase 3 there are probably perfectly good physical reasons why some
&gt; models might wave, e.g. increased evaporation -&gt; increased snowfall -&gt;
&gt; cooler northern summer.
&gt;
&gt;

I wouldn't mind hearing more about what we should expect. I have heard about adding 0.5 degrees K to the climate sensitivity due to variability. However I am unsure about this in respect of this model. Is most of this 0.5 variability due to EL Nino/La Nina and other oscilations that cannot occur with this slab ocean?

If we should expect 0.5 degrees variation in this model, why was the quality control for the nature paper set to be such a high standard as only allowing 0.02 degrees variation per year in phase 2?

Here is a CPView pic of Autumn 2057
<img src="http://cpdn.tuxie.org/crandles/303ica.pep1c10_5_147.JPG">

Not much of an effect, but I would think there is sufficient similarity in pattern to the cold equators to say that this is likely to be the same effect just weaker.
Visit BOINC WIKI for help

And join BOINC Synergy for all the news in one place.
ID: 9585 · Report as offensive     Reply Quote

Message boards : climateprediction.net Science : Another cold equator run

©2024 climateprediction.net