Flaw found on BBC app. Current models will be 'killed'

[NeoGen]

http://www.climateprediction.net/board/viewtopic.php?t=4697


I'm afraid to announce that we've recently discovered a major error in one of the files used by the climate model. The file in question specifies levels of man-made sulphate emissions but due to a problem with the file specification, models have been inputting greatly reduced levels throughout their runtime. The consequence of this is that aerosols responsible for "global dimming" (cooling) are not present in sufficient amounts and models have tended to warm up too quickly. The file specification error is also responsible for causing models to crash in 2013 which is how we originally came across the problem.

Unfortunately, all the data returned to us so far has been affected by this problem. While the data is scientifically very useful, and will certainly form the basis of future research (it allows us to investigate the full effect of greenhouse gas emissions without global dimming), it doesn't enable us to compare the models' performance against real world observations of the 20th century since such an important component is missing. In order to do the experiment we intended, we unfortunately have no choice but to start models again from the beginning.

So what now? We've spent the last few days fixing the error and we're in the process of generating and testing new model versions which should be ready in the next week or so. We very much hope you'll be prepared to keep running with us and, if so, all you need do is keep running your existing experiment and you will be automatically upgraded to a new model version next time you connect to our server via the automated system. We've removed all the old model versions so if you're waiting for a new workunit, they should be available within the week.

Apologies to all our participants on behalf of the CPDN team, it's a big disappointment to have to give you this news.
_________________
Nick Faull
Project Coordinator
Climateprediction.net
University of Oxford


EDIT: Damn quotes not working on the title... :?

[Evil-Dragon]

So I should just abort this workunit now and start with a new model? (this is for the BBC one I presume)

EDIT: I should really learn to read more...

[NeoGen]

I think right now their work queues are empty, so if you cancel the current one, you'll have no work from BBC to get.

If you're not running other projects, then you can as well keep the current model running until they "kill it", you'll get a few more thousands of credits in the meantime.

EDIT: By the way, no one actually needs to abort manually. All models will be "killed" remotely by the project server when time comes. I think it's the "trickle down" feature I've heard about but never saw in action.

[Evil-Dragon]

It's a bit of a waste really, i've done up to 1932 so far and now it's gotta start all again... they're gonna be really screwed for the BBC program in May.

"After a major screw up through lack of testing we have as much clue as you do when it comes to predicting what our climate will be like in the next 70 years time"

I bet someone is getting a real grilling from all this.

[NeoGen]

Just picked this up from the Q&A section of the project...

BBC Climate Change Experiment|Unrecoverable error for result hadcm3l_2lie_00121644_0 ( - exit code 99 (0x63))
If you see an exit code 99, you've just been killed! :lol:
http://bbc.cpdn.org/forum_thread.php?id=3170

[dAVE]

That's the code I just had but I went back to my backup so presumably I'll just have to wait to be killed again! Dave. :oops:

[Frederic Brillouet]

sh*t.

[NeoGen]

I wanna see what they'll say on the TV show they're supposed to be airing in about a month regarding the results of the project! :lol:

dAVE, you can kill it yourself if you want, instead of waiting for them to do it. At first I thought not, but I've checked and they've got full queues and are ready to send work already.
http://bbc.cpdn.org/server_status.php

[Strongbow]

There was I all happy that I finally had reached the future and about to change to 2008! :cry:

Not sure if I want to carry on, if they are going to scrap it all and make me start from scratch again!

I'll have to think about this! :?

[Strongbow]

Interesting I just got this...

16/04/2006 17:14:51|BBC Climate Change Experiment|Sending scheduler request to http://bbc.cpdn.org/bbc_cgi/cgi
16/04/2006 17:14:51|BBC Climate Change Experiment|Reason: To send trickle-up message
16/04/2006 17:14:51|BBC Climate Change Experiment|Note: not requesting new work or reporting results
16/04/2006 17:14:56|BBC Climate Change Experiment|Scheduler request to http://bbc.cpdn.org/bbc_cgi/cgi succeeded
16/04/2006 17:14:56|BBC Climate Change Experiment|handle_trickle_down failed: -116
16/04/2006 17:14:56|BBC Climate Change Experiment|handle_trickle_down failed: -116


Not sure if that is related!

[Strongbow]

It's started...

16/04/2006 17:17:26|BBC Climate Change Experiment|Scheduler request to http://bbc.cpdn.org/bbc_cgi/cgi succeeded
16/04/2006 17:25:27|BBC Climate Change Experiment|Unrecoverable error for result hadcm3l_47tn_00197217_0 ( - exit code 99 (0x63))
16/04/2006 17:25:27||request_reschedule_cpus: process exited
16/04/2006 17:25:27|BBC Climate Change Experiment|Computation for result hadcm3l_47tn_00197217_0 finished
16/04/2006 17:25:35|BBC Climate Change Experiment|Unrecoverable error for result hadcm3l_2ogc_00125458_1 ( - exit code 99 (0x63))
16/04/2006 17:25:35||request_reschedule_cpus: process exited
16/04/2006 17:25:35|BBC Climate Change Experiment|Computation for result hadcm3l_2ogc_00125458_1 finished


I'm back to 1920 on both cores!!! :cry: :cry: :cry:


Blimey! I read about it then it happens to me within 15mins! :-(

[drezha]

Not sure if this applies to the standard CPDN as well... :?
From a brief read of that post it would seem so because I have a had3clm but as they say it may be the subject of a future model so my guess is affect units on the standard one will carry one as normal but the BBC one's will terminate and restart...

[Dee]

Looks like I didn't go down without a fight! :P

16/04/2006 15:01:16|BBC Climate Change Experiment|Message from server: Project is temporarily shut down for maintenance
16/04/2006 15:01:16|BBC Climate Change Experiment|Project is down
16/04/2006 16:01:16|BBC Climate Change Experiment|Sending scheduler request to http://bbc.cpdn.org/bbc_cgi/cgi
16/04/2006 16:01:16|BBC Climate Change Experiment|Reason: To send trickle-up message
16/04/2006 16:01:16|BBC Climate Change Experiment|(not requesting new work or reporting completed tasks)
16/04/2006 16:01:21|BBC Climate Change Experiment|Scheduler request to http://bbc.cpdn.org/bbc_cgi/cgi succeeded
16/04/2006 16:01:21|BBC Climate Change Experiment|handle_trickle_down failed: null pointer
16/04/2006 16:01:21|BBC Climate Change Experiment|handle_trickle_down failed: null pointer
16/04/2006 16:01:21|BBC Climate Change Experiment|handle_trickle_down failed: null pointer
16/04/2006 16:17:19|BBC Climate Change Experiment|Unrecoverable error for result hadcm3l_r2ed_00327774_0 ( - exit code 99 (0x63))
16/04/2006 16:17:19|BBC Climate Change Experiment|Deferring scheduler requests for 1 minutes and 0 seconds
16/04/2006 16:17:19||Rescheduling CPU: application exited16/04/2006 16:17:19|BBC Climate Change Experiment|Computation for task hadcm3l_r2ed_00327774_0 finished
16/04/2006 16:18:22|BBC Climate Change Experiment|Sending scheduler request to http://bbc.cpdn.org/bbc_cgi/cgi
16/04/2006 16:18:22|BBC Climate Change Experiment|Reason: To fetch work
16/04/2006 16:18:22|BBC Climate Change Experiment|Requesting 8640 seconds of new work, and reporting 1 completed tasks
16/04/2006 16:18:27|BBC Climate Change Experiment|Scheduler request to http://bbc.cpdn.org/bbc_cgi/cgi succeeded
16/04/2006 16:18:27|BBC Climate Change Experiment|Message from server: Server can't open database
16/04/2006 16:18:27|BBC Climate Change Experiment|Project is down
16/04/2006 17:18:27|BBC Climate Change Experiment|Sending scheduler request to http://bbc.cpdn.org/bbc_cgi/cgi
16/04/2006 17:18:27|BBC Climate Change Experiment|Reason: To fetch work
16/04/2006 17:18:27|BBC Climate Change Experiment|Requesting 8640 seconds of new work, and reporting 1 completed tasks

[Evil-Dragon]

Back to 1921... again :\

[Majestic-12]

These things happen, nobody can be guaranteed from errors, what I think is good is that they admitted it honestly and did not try to hide it, errors are part of learning process, its okay so long as they are not repeated too often :)

[Frederic Brillouet]

Let's hope so! Imagine that the same thing happens with the new model, in 2050 or so. My sanity, my computer and those scientists would be severely endangered then :D

[Strongbow]

It's all over the news now...

Here's what the BBC had to say about it...

http://news.bbc.co.uk/1/hi/sci/tech/4923248.stm