http://www.climateprediction.net/new-experiment-launched-weatherhome-2015-western-us-drought/

We�ve just launched a new experiment looking at the causes of the ongoing drought in California, Oregon and Washington.

Has climate change made the drought more likely? Or is it the �blob�, the giant patch of unusually warm water off the West Coast in the northeast Pacific Ocean?

We�ve just put out over 20,000 models to study this phenomenon � you can read more about the drought in the Western US and how our experiment will help answer these questions on the weather@home pages on our website:

- weather@home 2015: Western US Drought
- Experiment setup
- Read more about the Western US Drought

We�ll be posting results up on our website as soon as we get the first models back, probably in about a week from now, so watch this space!

We�re doing this experiment with scientists based at Oregon State University � read their media advisory about the project.

Keep crunching!

Cheers,
Hannah

PS. We are aware that there isn't currently anything for Mac users to run right now, please stick with us until we release our next set of Mac simulations.

---

Hannah Rowlands
--
No longer Communications Officer for climateprediction.net, as of October 2015

Has climate change made the drought more likely? Or is it the �blob�, the giant patch of unusually warm water off the West Coast in the northeast Pacific
Ocean?

Or is the, "blob" itself a function of climate change?

Hi Dave - good question! My understanding is that climate change may both increase the magnitude of the sea surface temperature and the strength of the connection of this to climate elsewhere in the US.

The blob is not an entirely "natural" phenomenon, but neither is it "caused" by climate change - as always, the climate system is highly complex!

Our experiment isn't looking at the causes of the "blob" but should be able to say how much influence, separately, the "blob" and climate change have had on the current drought.

Cheers,
Hannah

---

Hannah Rowlands
--
No longer Communications Officer for climateprediction.net, as of October 2015

What will the designation be for the new project. Will it be a new type of hadam3p? Are tasks available yet? I don�t see anything new on the server status page.

---

What will the designation be for the new project. Will it be a new type of hadam3p? Are tasks available yet? I don�t see anything new on the server status page.

I think they're being sent out under the 'PNW' task type, which exists already.

I am running one now with BOINC as a non-service, but if anyone finds that they can run with BOINC installed as a service, I could add another machine.

To add to Richard's post:

These tasks (2013 _pxxx_ series) are 18 months rather than 12 months of earlier PNW tasks. Consequently, notice that estimated run time is ~50% longer than earlier PNW tasks.

---

"We have met the enemy and he is us." -- Pogo
Greetings from coastal Washington state, the scenic US Pacific Northwest.

http://www.climateprediction.net/weatherathome/western-us-drought/results/

We sent out the first models to our volunteers last week and so far we've had 607 returned to us - here are some plots using those initial models. Eventually, we will be running over 20,000 models, so we will add these to the plots as they come in.

The "climatology" results, that look at the possible effect of the "blob" on the drought, are not ready yet so we're just showing you the "actual" (world with climate change) and "natural" (world that might have been without climate change) experiments.

For an explanation of the 3 different sets of models we're running, read the Experimental Setup page.

For an explanation of what these plots are showing, read the Return Time Plots page.

You can see the full plots (so far) for precipitation, snow and temperature on the individual results pages for each State:

California
Oregon
Washington

We're going to be updating these plots regularly over the next few weeks, so do check back here as we go!

Cheers,
Hannah

---

Hannah Rowlands
--
No longer Communications Officer for climateprediction.net, as of October 2015

http://www.climateprediction.net/new-climatology-results-for-western-us-drought-experiment/

We now have the first results for our Climatological simulations, investigating the influence of removing the �blob� of warm sea surface temperatures off the western US coast.

The �blob� has a strong influence on the temperature, for example the climatological simulations without the �blob� are colder than the actual or natural simulations.

In the climatological simulations, it is interesting to see a different response in the precipitation between the different states. This is something our scientists will be investigating in more detail in the upcoming weeks.

Have a look at the results so far on the website:
http://www.climateprediction.net/weatherathome/western-us-drought/results/

Cheers,
Hannah

---

Hannah Rowlands
--
No longer Communications Officer for climateprediction.net, as of October 2015

Thank you Hannah for that information.

http://www.climateprediction.net/weatherhome-2015-western-us-drought-all-models-have-been-sent-out-about-half-have-come-back/

So, thanks to your fantastic efforts, all 22,000 models for this experiment have now been taken to run on your home computers, and nearly half have finished running and have been uploaded back to our servers.

The plots for the results so far have been updated to include all the models that have been returned to us � currently sitting at nearly 8,000 models!

Thanks again to everyone who has helped us with this experiment by running our models on your computer, we couldn�t do this without you.

There's 3 plots for each of the 3 states we're studying, but here's an image of the most recent results for snowfall in Oregon:


Cheers,
Hannah

PS. We've realised our mis-spelling on "celsius" and will correct it shortly!

---

Hannah Rowlands
--
No longer Communications Officer for climateprediction.net, as of October 2015

I am running one now with BOINC as a non-service, but if anyone finds that they can run with BOINC installed as a service, I could add another machine.

Boinc 7.0.36 can run all CPDN tasks as a service. Latter versions of Boinc are buggy!!, they CAN run: Hadam3p-EU, Hadam3p-ANZ, CAN'T run: Hadam3p-PNW, HadCM3-shot.

I am glad that you raised the subject. I have recently found that BOINC as a service also can't run 7.22 UK Met Office HadAM3P-HadRM3P Africa; at least the ones I have tried have failed, but then they run OK as a non-service, though the sample-set is not that large.

I am beginning to even suspect that some that usually run as a service may occasionally fail for that reason. But I have a bit of a problem dis-entangling that from the case that Folding on a GPU can cause problems also. To be safe, I now run CPDN only on a non-service install of BOINC, and don't do Folding on the GPU (though BOINC GPU projects such as POEM, Einstein and GPUGrid seem to be OK insofar as I can tell at the moment).

As 3rkko says, there is a very specific and identified incompatibility between BOINC clients later than 7.0.36 running in service mode, and the 7.22 / 7.24 CPDN applications for Windows deployed during 2014: yes, that includes HadAM3P-HadRM3P Africa. You can solve that any which way you like: not running service mode, not running a client later than 7.0.36, or not running the affected climate models.

This bug is exclusive to BOINC - any Folding issues are separate.

Nearly all the models have come back, time for the analysis.

It�s been a few weeks, but pretty much all the models have now come back to us, after being run on people�s home computers. As you can see on the plot below, we now have over 13,000 simulations to study, which is fantastic! We�ll keep collecting returned models, though, so please do keep crunching any you have on your computer.



Now we have most of the models back, we can begin the analysis, which will take several months.

Many thanks to all our volunteers for crunching all these models for us, it�s very much appreciated!

Cheers,
Hannah

---

Hannah Rowlands
--
No longer Communications Officer for climateprediction.net, as of October 2015

This bug is exclusive to BOINC - any Folding issues are separate.

The point was that Folding induces the problem in the HadCM3 shorts (and maybe others) when BOINC is run as a non-service, which it has to be for that project. But when I run BOINC as a service, I can Fold on the GPU without causing any errors to other type of work; all the errors are due to bad work units that fail on all machines. Whether that particular problem is due to BOINC or Folding is somewhat irrelevant; it is just an incompatibility.

As you may recall, the problem with Folding was mentioned here: http://climateapps2.oerc.ox.ac.uk/cpdnboinc/forum_thread.php?id=8091&nowrap=true#51995

So the point now is that I just avoid doing them on the same machine entirely, and the problem does not occur.