jump to navigation

Luminosity Profile June 19, 2007

Posted by gordonwatts in physics.

LumiProfileYesterday, Dmitri, one of our co-spokes person, showed this plot. I like this plot because it really made me rethink what is going on with our experiments luminosity. I like plots that jar me out of some preconceived notion I have.

Our experiment lives and dies by the amount of data it gets. Until recently the Tevatron accelerator was breaking high luminosity records. These high luminosities are tough to handle for the experiments: so much data arrives in such a short time that we often have to throw some of it on the floor just to keep up! Not good. We tend to focus most of our conversation on these issues.

The plot, however, shows that the bulk of the data we collect is otherwise. That plot is the luminosity profile for Run IIb. IIb is since we installed our inner layer of silicon — last summer. The x axis of that plot is the luminosity. The bulk of the events we have recorded were taking close to 50E30 — a fairly low luminosity (we get stressed out at about 200 or above). So it isn’t incorrect that our systems are still optimized for those lower luminosities: that is where we get most of our data!

That being said, however, there is a general rule of thumb in particle physics. New and interesting results and physics only occurs when you double your data. Most results coming out of the experiment are about 1 fb-1 of data. The next round of results will be between 2 and 3. After that we will need more than 4 to make real progress. In order to do that in a timely fashion (i.e. before the LHC starts up), the Tevatron will have to deliver more and more luminosity. One main way to do that is move the average of that plot to the right! And then we will need to re-optimize our system to deal with the higher luminosities!

This plot was done by Boris Tuchming for some of his muon studies. If you want to see it in context, check out D0’s talk at the P5 meeting.

Update: See comments for more details of the makeup of this sample.



1. a cornellian - June 20, 2007

What do the colors mean?

2. gordonwatts - June 20, 2007

Ah, yes. Right. So, the luminosity was divided up into 6 bins. So the bins are colored red, green, red, green, etc. That is the “bin” that the text on the plot refers to. The black line near the center of each bin is the mean of that bin. Finally, each bin is constant # of events, if I remember correctly. That is why the bins are more narrow near the peak, and broader out in the tails.

3. a cornellian - June 20, 2007


I clicked through the flickr link and read the same description. If that was there before i posted, i apologize.

Can I ask why 6 bins of equal size? I assume that tells you something interesting…but i can’t see what

4. gordonwatts - June 21, 2007

Actually — it wasn’t there until your comment. Then I figured I should put something on flickr as well! So, thanks for helping me to make things more clear!

I dont’ know why 6 bins were choosen. I think it was driven by something else that Boris was doing in his Muon study (perhaps he was classifying the muons by the average luminosity). I don’t think there is anything particular important in the choice for what I am discussing here (other than you don’t want to choose so many bins that you can’t really see the divisions). The binning information is also contained in the shape of the histogram — the lower the histo, the larger the bin — it goes by area.

I’ll ask Boris and post a follow up comment if/when he replies.

5. gordonwatts - June 21, 2007

Ok. I got some mail from Boris. I think the jist of what I said was mostly right, except for one thing.

– This plot comes from work trying to understand the acceptance of a muon as a function of luminsoity.
– Each bin (green or red band) has the same statistics, except the last one which is split into two.
– 6 bins were choosen to both provide as much splitting as possible and keep the statistics in the bin large enough to derive meaninful results.
– There are selection cuts that go into selecting this data sample. In particular, there is a trigger selection that is almost certianly coorelated with lumi — so the upper bins in lumi are probably underestimated by some amount.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: