Hi,
First time I see that.
The temperature and precipitation of the following model are falling in phase 2. It's on my stable machine and I had no error or problem with this wu.
Is it a bug of the graphic view or is it real ? What do you think of that ?
<a href="http://climateapps2.oucs.ox.ac.uk/cpdnboinc/result.php?resultid=429527">click me</a>

Hi Arnaud,

Have a look at this - my last one

<a href="http://climateapps2.oucs.ox.ac.uk/cpdnboinc/result.php?resultid=424234">424234</a>

Marj

---

_________________________________

&gt; Hi,
&gt; First time I see that.
&gt; The temperature and precipitation of the following model are falling in phase
&gt; 2. It's on my stable machine and I had no error or problem with this wu.
&gt; Is it a bug of the graphic view or is it real ? What do you think of that ?
&gt;

Hi,
I think you have got a model with unstable parameters. The cooling (or heating) in Phase 2 usually means that the ocean heat fluxes computed in Phase 1 (fixed ocean) are not exactly correct to hold the model in balance when it is running free in Phase 2. The Paper in Nature (and the data on the CPDN website) shows that quite a number of models show this to a greater or lesser degree. If it gets really bad (much worse than your example) then I think the science crew will ignore the result - unless someone gets a paper out of it of course ;-}

Andrew

---

Andrew

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

I expect you have a bit of a cold patch off the west coast of Ecuador. However there are quite a lot of models that show a really cold patch - under 6 degrees and perhaps even freezing.

The quality control test used for the first results paper was to reject models if there was a drift of more than 0.02K/Yr in the last 8 years of the control phase (2).

This is a tough test to pass:

2017 unique simulations were reduced to 1148 independent stable simulations by this quality control test.

The graph at <a href="http://www.livescience.com/environment/050206_new_iceberg.html">Second results</a> seemed reasonably consistent with a previous estimate made by UK Nick of perhaps 12% showing a cold equator. However the tough test applied has culled nearly half of the models.

---

Visit BOINC WIKI for help

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

&gt; Have a look at this - my last one <a href="http://climateapps2.oucs.ox.ac.uk/cpdnboinc/result.php?resultid=424234">424234</a>

Brrr, a cold one...

You may also want to read the following thread.

http://climateapps2.oucs.ox.ac.uk/cpdnboinc/forum_thread.php?id=1573

I have quoted from supplementary information about the cooling models.

---

Visit BOINC WIKI for help

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

Ok, thanks for the replies.
I thought that the freezing and cold equator problems occured at the beginning of Phase I and that the model stopped itself so I was surprised by the graph.
I didn't know it was so frequent: 12% is a high figure, IMO.

@Marj and Honza: Nice ones LOL :o)

Well, I have looked at the viz of the model, and it seems there's a cold patch west of equator: it's yellow, not orange...

Cheers...