Hello,
1.Chatting on a French Science Forum about the CPDN results, one person ask what are the parameters that cause such a hight raise of the temperature (he was talking of the 8-11K raise)?

2.The IPCC models predictions seemed to give a 3.5K raise around 2100, not 2050.
Can we deduce that CPDN results are more frightnening, alarmist than IPCC results ?
Why such a difference ? (sorry for my bad english, tonight)

---

Arnaud

Hi Arnaud,
i'm not sure about ad (1) but you will find answer to your (2) in Nature article.
Basically, one needs to take into account probabilty; there are only small portion of models showing >8K raise. IPCC and others (unlike CPDN) has apport of best fitting models, whereas CPDN is looking for ranges and plausible parameters.
Be aware that this is only my poor understanding and i might be wrong as well.

> Hello,
> 2.The IPCC models predictions seemed to give a 3.5K raise around 2100, not
> 2050.
> Can we deduce that CPDN results are more frightnening, alarmist than IPCC
> results ?
> Why such a difference ? (sorry for my bad english, tonight)
>
>

> Hello,
> 1.Chatting on a French Science Forum about the CPDN results, one person ask
> what are the parameters that cause such a hight raise of the temperature (he
> was talking of the 8-11K raise)?
>
On the phpBB forum there was a topic under the Extreme Climate section that had people posting model IDs for very warm runs. I had a couple that were about 9 C above "normal". Maybe looking through that forum when it comes back we could come up with some ideas on parameters that promote significant warming.

>1.Chatting on a French Science Forum about the CPDN results, one person ask >what are the parameters that cause such a hight raise of the temperature (he >was talking of the 8-11K raise)?


Parameters that appeared frequently in some of the hot runs reported (I only recorded 42)

Accretion constant 0.0004 30times (0.0001 12x 0.00005 0x)
Empirically adjusted cloud fraction 0.5 20times (0.7-0.6 17x 0.8-0.65 5x)
Ice fall speed 0.5 18times (1 17x 2 7x)
Critical relative humidity .95-.9 17times 0.95-0.7 17 times (0.95-0.6 8x)
Threshold precipitation over land 0.0001 18times (0.0002 14x 0.002 10x)

The last two or three of those could have occured by chance.

---

Visit BOINC WIKI for help

And join BOINC Synergy for all the news in one place.

From the Nature paper:

> If perturbations to another parameter (the entrainment coefficient)
> are omitted, the blue histogram in Fig. 2a is obtained, with no
> model versions >8K.

I think the suplemental material had something about parameter values but can't find the URL at the moment.

---

____________________________<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>.

Hmm don't trust my analysis: too many runs that are barely warmer than average.

Honza did have a classic run which was very hot (roughtly 9½K?) sensitivity and had the default entrainment coefficient.
http://cpdn.comlab.ox.ac.uk/user/getrun.php?stype=AT&amp;runid=167436

so maybe they havent included enough runs to let those histograms fully settle down yet.

The paper says
"We can illustrate the importance of parameter choices by subsampling the model versions. If all perturbations to one parameter (..) are omitted, the (red) histogram in fig. 2a is obtained."

I wonder why they omitted the perturbations rather than presenting three different histograms for each of the three values of each parameter? Surely that would illustrate the importance more clearly?

---

Visit BOINC WIKI for help

And join BOINC Synergy for all the news in one place.

&gt; I wonder why they omitted the perturbations rather than presenting three
&gt; different histograms for each of the three values of each parameter? Surely
&gt; that would illustrate the importance more clearly?
&gt;

Given Natures prestige position as the journal with the highest "impact factor" in the world, space is at a premium. Adding more figures to the Letter would have run over their notional space budget (ca 3 pages??). So more pictures less explanation (and less chance of getting it published!!!)

I look forward to the extended version of the paper containing the supplemental material that Dave Frame mentioned might become available to CPDN members.

Andrew

---

Andrew

<a href="http://cpdnforum.info">CPDNforum<a>

I remember somewhere in the Nature article it said the most sensitive parameters were the ones that had to do with cloud formation.

---