Restart required on the BBC Climate Change Experiment

[Heather T]

Where do I find the scrolling message? On the globe? in the Help? Messages? I find it quite difficult to find my way around the site.

[Stratcat®]

I'm just glad I'm not the guy that gave Carl or Tulo the bad data file!

Strat - Suspecting somebody's in big trouble

[MikeMarsUK]

The scrolling message is on the globe, but it may take a trickle or two before it appears.

[Rory]

should i delete the old model and just leve the new model running, i only got to 1930 but i work funny hours, now iam going to have to run 24/7 to regain the hours lost!!

[MikeMarsUK]

Yes, 'abort' the model if it's a hadcm3l_ or hadcm3lb_ (hadcm3ln_ and hadcm3lbm_ are good)

[tony peduzzi]

Experiment reset after more than 260 hours processing.

I've uninstalled.

[crandles]

Experiment reset after more than 260 hours processing.

I've uninstalled.

I think that is a shame. Two months lost work would be bad but not disasterous, and it is not even that bad as the work is still useful in a different unintended way. However if it leads to lots of people disappearing that seems to be the best way to make a bad situation worse.

I hope you reconsider and acept that the course of science never flow in some straight path but u-turns are often needed. It is of course entirely up to you.

[madbiologist]

Experiment reset after more than 260 hours processing.

I've uninstalled.

Please don't detach from CPDN. Due to the regular trickles the model sends every model month, and the upload every 10 model years, Oxford University has the results of your 260 hours of processing, and as stated in the post which started this thread http://www.climateprediction.net/board/ ... sc&start=0 your results will still be very useful scientifically.

You will also retain the credit for the work you have done.

[Son Goku]

Most software gets its full test from the users after publication - the authors of test scripts can never cover all the eventualities. That is why you continually get updates. Also the problem may go back to the definition of the global-dimming conditions.

It's also the reason many projects sit in beta for awhile (such as Windows Vista right now), even after it has been in-house tested. And sometimes, even then... I don't want to contemplate when Windows ME was in beta... I was an MSDN pro subscriber at the time, so had access to it from RC 1. And to make a long story short, many MSDN subscribers wrote putting Millenium Edition on their development computers, even though they payed nothing additional for the CD...

To suggest that the person responsible for the error be sacked as a matter of principle would mean that Microsoft, IBM and other major software suppliers would be very short of staff in their development labs - and in need of rapidly revolving doors at their entrances. At least HR would be well resourced!

One thing would be certain. The entire Internet Explorer development team would have been sacked a long time ago. Not just bugs, but security issues. In fact, almost everytime something's on Windows Update, there's a fix for IE, and almost all of those tend to deal with potential security vulnerabilities and exploits.

tbh, I was taken quite by surprise when the model crashed. I didn't even know a mechanism was in place where a model could be aborted from the network. Makes sense in a way, as a fail safe; but to say I was scratching my head a bit when it failed on just loading a model, that got beyond where it was when I suspended it (to let lhc crunch, when it got work again), would be to put it mildly. Couldn't figure out why my efforts at a WU restore weren't going, or what to do with the core client that wouldn't let me restore...

Course I also never did get a successful sulpher run on this computer, but then coupled seemed to be going fine (along with seasonal attribution); which suggested that whatever issue I was having might have gotten worked out with this version. And then

In a way though, as much as it all hit me by surprise, and I watched my 100 hours crunch time disappear; it's also good to know the people running the experiment have displayed a certain level of honesty in pulling these models. Too often have I been in discussion with some of the more, shall I say quite political doomsday theorists. And when I've responded to some of their claims, with some actual climate data that doesn't tend to suggest such an alarmist take, or suggested the climate is a more intricate system with it's own feedback mechanisms; umm they have gotten more political, and we end up at logger heads...

Anyhow, as much as it can take one aback to lose all this crunch time, it's also good to know that there was a willingness to pull these models and set the record straight. In the long run, such willingness can leave one a bit more confident with the results we'll end up generating, then if the matter were just allowed to continue as such...

[GreatInca]

Is CPDN and BBC identical? I joined BBC through boinc. I barely have any credits in BBC as of yet. My CPDN is still doing a sulphur_isws* workunit. My BBC is doing HADCM3l.* My CPDN seasonal is HADAM3h_n_*

Most of my CPDN workunits error about 200-400 hours into the workunit anyway (my work computer usually last longer).

My resource shares for the CPDN projects are CPDN vanilla 10%, CPDN seasonal 10%, CPDN BBC 5%. SETI is 30%, Einstein is 20%, LHC 15%, QMC 5%, HashClash 2.5%, ufluids 2.5%. I don't expect my CPDN projects to complete by the dealine and the CPDN workunits keep my BOINC in NDF mode which assists in me pigging as many of LHC's limited supply of WUs as possible (and it crunches LHC first because of LHCs short deadlines).

[geoff4711]

I discovered my world had reset to 1921 yesterday and spent hours getting it back to 1945 via my backup in briefcase.(Thought it had crased) now I'm not sure if I am running the latest one or not.
Is there a version No. or similar to identify the correct variant?
by the way there were 2 folders marked "climate change" inside the main one in programs.Is there any way of telling which one to update from briefcase when it does crash?

[MikeMarsUK]

The bad ones will be titled:

had3cml_ ... or had3cmlb_ ...

The good ones will be:

had3cmln_... or had3cmlbm_...

[herme3]

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.

Could you explain more about what this means? If my model was in the late 1940's when it was reset, did anything that my computer processed help? Was data being uploaded in the trickle messages? What exactly will this data be used for? How can anyone can accurately predict global warming without global dimming data? Wouldn't it be like trying to predict the weather if you left out clouds?

[grahamd]

Pull yourself together Herme3 or take a holiday. Read a few posts. It means exactly what it reads.
Try and write something positive.

[richie]

My new bbc climate model ends with an "n"

That it to say it is named "hadcm3ln"

Does this mean I am now running a new version?

This new work package was downloaded today 17th April 2006. I'd hate to think this is another old version. For those in the know please confirm if this is ok to use.

Cheers

Rich