anarchist_nomad (![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png)
anarchist_nomad) wrote2007-04-08 05:28 pm
anarchist_nomad) wrote2007-04-08 05:28 pm
Entry tags:
What Do You Do With a Doctorate In Physics? What Is My Life Going To Be?[*]
A couple of weeks ago, ![[livejournal.com profile]](https://www.dreamwidth.org/img/external/lj-userinfo.gif)
xirpha asked me, via LJ comment, the following question:
May I be so bold to ask as what is your carrier goals? You seem to have spent all of your time that I have known you working as a technician than a scientist. I understand than research is mostly grunt work, but do you hope to run your own research project?
Rather than reply via comment, I thought about my answer... then decided it would be better suited to its own post. A chance tobore my friends to death explain just what it is that experimental particle physicists do. By the way, if any other particle physicists who read my journal want to chime in to add or clarify, please do. I'm talking to you, gyades and madandrew!
So, what do we do? In short, a bit of everything. That's not precisely true, of course, but it is not a bad first approximation. To be a good experimental particle physicist, you need to know a good deal physics... and a bunch of maths[**]... and a bit of computer programming... and a fair knowledge of electronics hardware... and computer hardware... and data analysis skills. It also helps to have good writing skills and good presentation skills. Basically, to do good research in my chosen sub-field, you need to be something of a jack of all trades.
Mind you, you don't need to know enough programming to be a code monkey -- though there are definitely people who do! Likewise, you don't need know know enough maths to be a mathematician, nor enough electronics to be an electrical engineer. What you do need is a basic proficiency in all these areas (and more than just basic in the physics, of course!). There are definitely people who have stronger abilities in one area or the other on this list. So you get folks who are predominantly software people, or predominantly hardware people. But it is important to keep a balance! I have seen people who become too enmeshed in the hardware have difficulty advancing their careers, because there is not enough physics output.
I can completely see where the comment comes from about my work seeming more like that of a technician than a scientist. Of late, I have spent a lot of time calibrating thermometers, cooling down dilution refrigerators, and whatnot. On Pierre Auger, I did a small bit of analysis... but most of my time was spent constructing and taking telescope calibration data with a self-contained laser facility. On Auger, that was a bit of a mistake on my part -- I did become too wound up in hardware-only activities. I have had concerns about the same thing happening on CRESST, and have discussed this with my boss. He pointed out that, to advance my career, I need publications that I can point to as being work that I played a major role in. The CRESST data will be analyzed and published -- with my name on the author list -- from the work that I have already done for the experiment. Thus, he feels that I should concentrate on other tasks -- which I cannot, in good conscience, write about here... but may yet get to that private e-mail about them. To get to those tasks, I have had to solve the problems with the Kelvinox, and I have to sort our electronics and calibrate our thermometers. And when those tasks are done, they should lead to publications. If it plays out this way, it is a healthy progression.
xirpha's question is also natural because he did not know me when I was a graduate student on the Super-Kamiokande experiment. Perhaps my time on Super-K is a better example of how the whole process works:
When I first arrived at SK, I started off doing hardware calibration tasks, like using a radioactive NiCf source to calibrate the detector, and using an electron LINAC (LINear ACcelerator) to do a much better calibration of the detector. As time passed, I also took on software calibration tasks, like measuring the speed of light in Super-K water to an accuracy of 0.7%, monitoring the water transparency on a weekly basis, and monitoring the energy stability of the detector on a semi-annual basis. I also moved on to data analysis tasks, like designing new selection criteria to remove backgrounds and designing a fitter to reconstruct the opening angle of Cherenkov light cones. True, there were also more grunt tasks, like being responsible for an off-line data backup facility (which involved hardware, software, and managerial skills -- not to mention far too much time), and assisting with the upgrade of the detector in 2001 as well as the rebuilding in 2002. Some of the grunt tasks may seem like they had nothing to do with physics (e.g., getting a mine driver's license for Japan).
In the end, however, my single greatest contribution to the experiment -- and to my career to date -- was my dissertation analysis: A search for supernova relic neutrinos (SRN). Though I did not find any, my results were two orders of magnitude (about a factor 100) better than any previous searches... and they were able to rule out some theoretical models. Perhaps more importantly, my results are only a factor of three away from being able to either find the SRN -- opening an age of neutrino cosmology -- or ruling out all models. A factor of three might sound like a long way to go, but consider that I already gained a factor of one hundred over previous experiments. Based on my results, new ideas have sprouted for how that last factor of three can be obtained.
Still, the work was not done with the production of a physics result. Besides simply writing up the analysis and results in my dissertation, I also wrote a paper that was published in Physical Review Letters -- arguably the most august journal in physics -- and was granted first authorship on the paper (out of a collaboration of 125 authors). Said paper has been cited, to date, seventy-two times since it was put on the pre-print server in September 2002. That is seventy-two citations in fifty-five months, or an average of one citation every twenty-three days. Not too shabby! And, of course, in 2002 and 2003, I gave seminars and conference talks on my result.
So there is a picture, from start to finish, of my time on a particular experiment. I did hardware work, calibration work, software work, tonnes of grunt work, data analysis, and -- finally -- produced a physics result. Which was then written up, as well as presented in person, to the scientific community. Of course, I spend five years on Super-Kamiokande, as opposed to only two and a half on Auger and only one (so far) on CRESST. So, while it may seem like I am being more of a "technician" at the moment, this work is actually a part of the research physicist's job... and does eventually lead to producing science.
Clearer now? Any questions? Remember -- there will be a quiz next week!
[*] Eleven years of college, and plenty of knowledge have earned me this useful(?) degree...
[**] I am switching to the British shortening of mathematics.
xirpha asked me, via LJ comment, the following question:May I be so bold to ask as what is your carrier goals? You seem to have spent all of your time that I have known you working as a technician than a scientist. I understand than research is mostly grunt work, but do you hope to run your own research project?
Rather than reply via comment, I thought about my answer... then decided it would be better suited to its own post. A chance to
gyades and madandrew!So, what do we do? In short, a bit of everything. That's not precisely true, of course, but it is not a bad first approximation. To be a good experimental particle physicist, you need to know a good deal physics... and a bunch of maths[**]... and a bit of computer programming... and a fair knowledge of electronics hardware... and computer hardware... and data analysis skills. It also helps to have good writing skills and good presentation skills. Basically, to do good research in my chosen sub-field, you need to be something of a jack of all trades.
Mind you, you don't need to know enough programming to be a code monkey -- though there are definitely people who do! Likewise, you don't need know know enough maths to be a mathematician, nor enough electronics to be an electrical engineer. What you do need is a basic proficiency in all these areas (and more than just basic in the physics, of course!). There are definitely people who have stronger abilities in one area or the other on this list. So you get folks who are predominantly software people, or predominantly hardware people. But it is important to keep a balance! I have seen people who become too enmeshed in the hardware have difficulty advancing their careers, because there is not enough physics output.
I can completely see where the comment comes from about my work seeming more like that of a technician than a scientist. Of late, I have spent a lot of time calibrating thermometers, cooling down dilution refrigerators, and whatnot. On Pierre Auger, I did a small bit of analysis... but most of my time was spent constructing and taking telescope calibration data with a self-contained laser facility. On Auger, that was a bit of a mistake on my part -- I did become too wound up in hardware-only activities. I have had concerns about the same thing happening on CRESST, and have discussed this with my boss. He pointed out that, to advance my career, I need publications that I can point to as being work that I played a major role in. The CRESST data will be analyzed and published -- with my name on the author list -- from the work that I have already done for the experiment. Thus, he feels that I should concentrate on other tasks -- which I cannot, in good conscience, write about here... but may yet get to that private e-mail about them. To get to those tasks, I have had to solve the problems with the Kelvinox, and I have to sort our electronics and calibrate our thermometers. And when those tasks are done, they should lead to publications. If it plays out this way, it is a healthy progression.
xirpha's question is also natural because he did not know me when I was a graduate student on the Super-Kamiokande experiment. Perhaps my time on Super-K is a better example of how the whole process works:When I first arrived at SK, I started off doing hardware calibration tasks, like using a radioactive NiCf source to calibrate the detector, and using an electron LINAC (LINear ACcelerator) to do a much better calibration of the detector. As time passed, I also took on software calibration tasks, like measuring the speed of light in Super-K water to an accuracy of 0.7%, monitoring the water transparency on a weekly basis, and monitoring the energy stability of the detector on a semi-annual basis. I also moved on to data analysis tasks, like designing new selection criteria to remove backgrounds and designing a fitter to reconstruct the opening angle of Cherenkov light cones. True, there were also more grunt tasks, like being responsible for an off-line data backup facility (which involved hardware, software, and managerial skills -- not to mention far too much time), and assisting with the upgrade of the detector in 2001 as well as the rebuilding in 2002. Some of the grunt tasks may seem like they had nothing to do with physics (e.g., getting a mine driver's license for Japan).
In the end, however, my single greatest contribution to the experiment -- and to my career to date -- was my dissertation analysis: A search for supernova relic neutrinos (SRN). Though I did not find any, my results were two orders of magnitude (about a factor 100) better than any previous searches... and they were able to rule out some theoretical models. Perhaps more importantly, my results are only a factor of three away from being able to either find the SRN -- opening an age of neutrino cosmology -- or ruling out all models. A factor of three might sound like a long way to go, but consider that I already gained a factor of one hundred over previous experiments. Based on my results, new ideas have sprouted for how that last factor of three can be obtained.
Still, the work was not done with the production of a physics result. Besides simply writing up the analysis and results in my dissertation, I also wrote a paper that was published in Physical Review Letters -- arguably the most august journal in physics -- and was granted first authorship on the paper (out of a collaboration of 125 authors). Said paper has been cited, to date, seventy-two times since it was put on the pre-print server in September 2002. That is seventy-two citations in fifty-five months, or an average of one citation every twenty-three days. Not too shabby! And, of course, in 2002 and 2003, I gave seminars and conference talks on my result.
So there is a picture, from start to finish, of my time on a particular experiment. I did hardware work, calibration work, software work, tonnes of grunt work, data analysis, and -- finally -- produced a physics result. Which was then written up, as well as presented in person, to the scientific community. Of course, I spend five years on Super-Kamiokande, as opposed to only two and a half on Auger and only one (so far) on CRESST. So, while it may seem like I am being more of a "technician" at the moment, this work is actually a part of the research physicist's job... and does eventually lead to producing science.
Clearer now? Any questions? Remember -- there will be a quiz next week!
[*] Eleven years of college, and plenty of knowledge have earned me this useful(?) degree...
[**] I am switching to the British shortening of mathematics.