jump to navigation

Will it really take ATLAS 3 years to see 5 sigma Higgs? August 20, 2008

Posted by gordonwatts in ATLAS, CERN, D0, Fermilab.

Probably (for ATLAS new predictions on this should be released in a few months). But in the context of the Tevatron and the LHC Higgs search that isn’t really what is important.

The ATLAS prediction that it might take 3 years to reach the 5 sigma level for a low mass Higgs discovery got a lot of airplay. It got me to thinking. Lets say the two accelerators are in close competition for the Higgs. The Tevatron can really only speak to the 3 sigma level. It isn’t ever going to get to the 5 sigma level. Further, at the Tevatron the CDF and DZERO experiments will have to combine their results to even reach this 3 sigma level. So, I find it highly unlikely that the LHC will sit back and let the Tevatron get away with this. I certainly wouldn’t (and I’m on a LHC experiment). So what to do? Obvious – beat the Tevatron at its own game: combine results from CMS and ATLAS and the 3 sigma level will be obtained much more quickly. At that point the LHC has stolen the thunder from the Tevatron and CMS and ATLAS can now race each other to individual discoveries of the Higgs at the 5 sigma level.

I don’t expect the experiments to combine for the 5 sigma discovery (I could well be wrong, of course – I know of no plans to not do this or to do this!). There are many forces at play that are driving each experiment to make the first paper submission of a 5 sigma signal. This may, indeed, be what gives the Tevatron space to slip in with a 3 sigma evidence paper. And in the grand scheme of things – the Tevatron goes out with a 3 sigma evidence and the LHC with a 5 sigma discovery – that doesn’t seem like a bad “split”. But who has ever heard of the free market working like that!?

As a member of DZERO I want to push as hard as possible to nail a low mass Higgs. As a member of ATLAS, I want the experiment to scramble as fast as possible to get the Higgs – evidence and discovery. After all, that is one of the LHC’s main points.


DZERO Workshop Redux August 18, 2008

Posted by gordonwatts in D0, life, travel.
add a comment

IMG_4573Well, if there is one thing to be said about the DZERO workshop, it is that the recent breaking of the Higgs SM limit line by CDF and DZERO has certainly injected excitement into the experiment. On the other hand, the LHC startup is right around the corner (officially, Sept. 10). Many of the talks given were infused with both of these themes. The theme of the workshop isn’t hard to guess: the Tevatron is schedule to give another 4 fb-1 of data or so — doubling our current dataset — is the experiment ready to absorb that? Especially as people drain away from the Tevatron experiments to the LHC experiments.

On one hand, many of the future talks would split the physics they discussed into physics that only the Tevatron could do vs physics that both the Tevatron and the LHC could do. The implication being, of course, that the LHC would quickly produce better results once it was up and running.

For the Higgs searches, it was the opposite — there is a small window of opportunity that must be grabbed while the grabbing is good. Indeed, I think it is safe to say that most of the experiment is focused on getting the Higgs signal.

At the last minute I was asked to give a talk on my daqAI work. This is artificial intelligence for the data acquisition system at DZERO. Actually, there is nothing “AI” about it – it is just a bunch of if-then-else statements that recognize a problem and either tell the shifter or attempt to fix it themselves. The up shot is better up time. Actually, DZERO’s data taking efficiency is amazing. Above 90%. When we hit 80% we claim we’ve had a bad week. Our competition, CDF, is not doing so well. While I’d like to think that daqAI helps towards this (and I suppose it does a bit) the real reason is the people we have working on this. CDF – if the Tevatron is going to say anything about the Higgs then you guys need every drop of data… Get to work! πŸ˜‰

Another really interesting session I went to was the b-tagging one. I used to co-run the b-tagging group at DZERO. It was especially satisfying to see some large projects that I attempted to get going while I was the co-convener actually completed. They weren’t completed because they continued using myself – people mostly started from scratch – bit still… And the fact that ex-students of mine had a hand in getting those things done. Very cool! And, of course, there is all the ongoing work – some of it rather interesting.

Lots of other interesting stuff, but I can’t talk about it of course.

Finally, our Prague hosts did a great job. That picture above is Julia dancing to traditional Czech music. πŸ™‚

I’m on vacation now, in Cyprus. My wife was full time Mom last week while I was attending the DZERO workshop. This week it is her turn – she gets to keynote at the physics education conference here in Cyprus while I’m a full time Dad. After that we will be busy moving back to the USA. In this hotel they charge 20 euros a day for internet… so I’m not sure if my blog will be updated much! But we’ll see…

Play it safe, or… August 17, 2008

Posted by gordonwatts in ATLAS, CERN, D0, Fermilab.
add a comment

There was an unspoken theme at the DZERO workshop this week. Stick with the Tevatron for a huge, but iffy, payoff. Or switch to the LHC now because it is a “sure” bet (as sure as anything gets in research).

This is all about the Standard Model Higgs search at the two accelerators. If such a Higgs does exist the LHC is bound to discover it. The LHC has some “difficulty” at low mass Higgs (below about 125 or so). Difficulty for the LHC means it could take up to 3 years for a single experiment to declare a 5 sigma discovery, the gold standard of “discovery”.

At the Tevatron the Higgs analysis is all about difficulty. Each new Higgs result you hear or read about is a tour-de-force of new techniques and new methods of extracting every last bit of signal out of the experiments. As a graduate student I never remember techniques this sophisticated. And the LHC pre-trial analyses are not as sophisticated either (on the other hand, they don’t need to be).

Global fits to the Standard Model currently predict the Higgs to be low mass – between 114 GeV and 120 or 125 GeV. The Tevatron is currently x2 away from being sensitive to this mass range. By doubling our dataset to 6 fb-1 of data and making a number of improvements to our analyses, we expect that we should be there. These improvements are not easy – it will require a lot of work and a lot of people. Nor are they assured. At best, if the Higgs is there, and we aren’t unlucky, we should be able to see it at the 3 sigma level. But never the 5 sigma discovery level. That will have to be left to the LHC in any case.

So is it worth sticking with the Tevatron? Well… the payoff would be huge to see something at the 3 sigma level. So it is like a lottery with high stakes. The chance of winning is not all that sure, but the jackpot is big!

Me? Well, I’m working on both the LHC and the Tevatron (as are many US physicists). I have a student working on the Higgs search at Fermilab, for example. I’m deeply involved in a number of topics at the LHC as well.

What will happen? Hard to tell. Things to watch? Well, that is easy. There are only two things that really matter here – the performance of the Tevatron and the performance of the LHC. Each physicist who is on both collaborations is performing some complex calculus to optimize their time on the two experiments depending on the chances of success.

I wish us all luck. πŸ™‚

I’m Not In France Anymore! August 10, 2008

Posted by gordonwatts in D0, physics life, travel.

IMG_4250That is Pork’s Knee (or Pork’s knuckle as it was labeled on the menu). And it was good. Big enough to search two, and J-mo ate a little too. Somehow I don’t think you would ever find a dish quite like this in France. And the beer along with it – I’ve not tasted beer that good in months! I love French food and wine, but there are some things done better elsewhere…

I’m in Prague this week, along with the family, attending D0’s yearly workshop. I don’t know how much posting will occur (probably not much). And the week after I’m on vacation being a Dad and Paula attends a big conference. I’m pretty sure almost not posting will occur then.

In the mean time I’m going to enjoy as much Czech beer as I can!

How Hard Will The Hunt Be? August 6, 2008

Posted by gordonwatts in D0, Fermilab, Higgs, physics.

Yesterday I mentioned that the Tevatron experiments had finally started to rule out the Higgs. I thought I’d post another plot that shows exactly how hard it will be – and so gives you an idea of how much hope the Tevatron has of actually catching the Higgs. Click on the plot to get an enlarged version of the jpeg (here for details).

The most important lines in that plot are the black one (1-CLs Observed) and and the 95% CL thick blue line. The thick blue line is the point at which, in our best statistical estimate, we are 95% confident that we have not observed anything. While the blue line is the “goal”, the black line is where we are now – the current observation. A lot goes into that black line – many different physics analysis contribute (from both D0 and CDF), the physics of the Higgs decay, the physics of how the Higgs boson is supposedly made, and how good our detector is at seeing the Higgs. As you can see, we have just peaked above the 95% level near 170. And that is what allows us to say that we’ve excluded the Higgs around 170 GeV.

Now, the future. You’ll note that the curve is pretty flat near where it peaks above 170. That says to me that when we add more data and minor analysis improvements we will be able to quickly broaden the amount of the observed line is above the 95% CL line. Where the black line is steeply falling, however, it require a huge amount of work (even if it is possible at the Tevatron).

Finally, in yesterday’s post the plot started at 114 GeV. This one starts at 155. What about everything from 114 to 155? Yes — we are working on that. For example, at D0 we have individual results already (and if you look at this plot, given the discussion, you can see that how we are doing as far as getting towards ruling things out at low mass – though the plot is a very different type of plot – but you can guess what is going on if you are not familiar with it). I couldn’t find the recent update of the CDF combined results. But the low mass combination between the experiments was not completed in time for ICHEP. I’m hopeful that we will see it soon – but as they say, it ain’t out until it is ready to be out!

A hunting we will go… August 5, 2008

Posted by gordonwatts in D0, Higgs, physics.

See that little red blob around 170? That is the Tevatron starting to seriously tackle the final big physics problem left on its plate. Where is the Higgs? The question is — will it finish the job before the LHC starts producing real physics?

The numbers on that plot are the mass of Higgs boson, the final bit of the Standard Model we physicists haven’t directly observed. The last experiment to search for the Higgs were the LEP experiments. As you can see, they searched up to 114 GeV. The Tevatron is searching from 114 up as high as it can go — it so happens the first bit it was able to exclude was around 170 GeV in mass.

The Higgs mechanism is what gives most particles mass. If it was absent from our theory then many masses (and other things) we have already measured would be wrong. That does not mean, by the way, that the Higgs has to exist – but something like it does have to exist. The Standard Model Higgs is just the simplest explanation that we came up with fix the masses. If that whole range is searched and nothing is found – that would be huge news. And very puzzling!

Press Release Here. And combined CDF and D0 note describing the analysis here.

ICHEP Should Be Good! July 31, 2008

Posted by gordonwatts in Conference, D0.

ICHEP is really getting under way this Sunday — it should be a good conference – with a bunch of interesting results. Some I know about, some I’ve heard rumors about — I’m eager to see what is actually going to make it out there. The only bummer for me is they aren’t using CERN’s agenda system so I can’t just DeepZoom the thing!

Oh — an D0 just passed another milestone in data it has collected — over 4 fb-1 now! The results shown at ICHEP are on the 3 fb-1, however.

Bye & Good Luck, Aran July 9, 2008

Posted by gordonwatts in D0, University of Washington.

Celabrating...Well, it’s happened. Aran has been with the University of Washington for the last 5 years – as a post-doc. And as of July 1st, he has moved on. Without him UW’s presence at D0 would be much less. He has been UW’s public face there – maintaining the Level 3 DAQ system, and running the single top group during our evidence paper. He has also been one of the best student mentors I’ve ever seen.

But fear not — he has definitely not left particle physics — he is now a professor at the University of Rochester!! I find this particularly cool because that is where I went for graduate school myself.

Best of luck!!

Like Red Meat To Dogs June 7, 2008

Posted by gordonwatts in D0.

So, what is the quickest way to fill up an experimental particle physicist’s InBox? Write an email that looks something like this:

I’d like to propose we change our shift schedule. Instead of shifts from 12pm-8am, 8am-4pm, and 4pm-12pm how about 4am-12am, 12am-8pm, and 8pm-4am? I think it would help everyone get a better night’s sleep.

Whew. That was proposed on one of D0’s mailing lists yesterday and I think I’ve got over 15 massages on that topic alone in an hour and they are still flowing in.

It is a great debate. I personally don’t care one way or the other. It is serious stuff for some — especially when families and teaching schedules are involved. And of course, everyone’s biological clock will likely make them prefer one or the other.

Good luck resolving this, D0. I’ll just continue to do the owl shift (12pm-8am) or the vampire shift (8pm-4am) if we change the schedule.

New Level 3 Lesson May 22, 2008

Posted by gordonwatts in computers, D0, Trigger.
add a comment

Ok, here is a dumb lesson I’ve learned the hard way. And thanks to many others who helped resolve it. Lets say you design a distributed system – like your online and trigger and data collection system at D0. This is a medium sized system — perhaps 500 boxes and several 1000 CPU’s at this point. It is key to note that this is a heterogeneous system — many of those boxes are doing different things and have to be custom configured.

Now, since it is heterogeneous, but a distributed system, and all the boxes have to communicate with each other, they have to have a way of finding each other. You definitely can’t use raw DNS and the machine name. Computers change. Sometimes you want to do a hot-swap to an experimental system. Your DNS is managed by a central facility so the turn-around can be a day – and when the accelerator is delivery beam you need less than an hour.

So you have to decide on some sort of name service. Some service that can take a name and reply with a machine. If it is done right, this will disappear into the infrastructure and you’ll not even be aware it is there after a few years.


Lets see, we’ve been running since 2001. In about 2003 we started using what the “sanctioned” name server for our Level 3 Trigger and DAQ part of the system. Of course, you have to make sure you know where that name server is for all this to work. We had an alias in DNS for that purpose.

And it turns out that our stuff is one of the few things left using that nameserver. Everyone else loads a python file on the command line. I’d originally designed our system so that you could change the location of a system on the fly without having to reboot one of the components – so the python approach was never considered. And the online system recently cleared out a bunch of machines.

The name server was moved. And that alias? Well, everyone forgot and so it wasn’t established. And then slowly, over time, parts of Level 3 started to fail. Thank goodness it was the monitoring code that failed first. But there were several hours of panic. All of us had forgotten how the system works it has been so long.

Maintaining the same system running for years is so weird. Almost all the code I write I think “Ok — get it running, debugged, and check it in and move on.” Keeping some of it running for years, however, there are other considerations. I bet there are whole books on this. Too bad we HEP people never take the time to read that sort of thing before we do our software development…