In this three-part blog post, I want to share with you how I've used Cloudant to manage my data while working on fundamental particle astrophysics. In this first post, I'll tell you a little about the really cool physics I have worked on and I'll save the database part for later. Perhaps at the end of this I'll even tell you about an old school NoSQL (PreSQL?), Fortran/ZEBRA-based database from hell that was built and used by my colleagues during my Ph.D. work. Seriously, you have no idea!
Before I joined Cloudant last month, I was fortunate enough to spend the last 15 years of my life studying the fundamental interactions in our Universe. I helped solve the Solar Neutrino Problem working with the SNO experiment while at the University of Washington. Most recently I was searching for dark matter particles as a member of the EDELWEISS collaboration while at the Institute for Experimental Nuclear Physics at the Karlsruhe Institute of Technology in Germany. On both of these experiments, I've worked with the most brilliant scientists and engineers from around the world. I'm very proud and humbled to have been a part of these magnificent endeavours. Now I've moved on to another fascinating project: Cloudant - a company filled with very talented individuals with whom I share a passion for all things data!
These blog posts will be about how, over the past two years, I managed EDELWEISS data with Cloudant.
While there are a number of robust observations that provide very good evidence for the existence of the dark matter, perhaps the most compelling is of galactic rotation curves, which is the phenomenon that the speed of stars at the edges of galaxies are significantly greater than the speed expected by Newtonian physics. The expectation is that the speed will decrease with distance from the center (speed ~ sqrt(1/distance)). Remarkably, and invariably, they don't! The speed as a function of distance remains relatively flat. This is completely inconsistent with the observation that most of the matter is at the center of the galaxy (the luminous matter), and consistent with some unobserved matter in the galaxy. Voilà - dark matter!
EDELWEISS looks for dark matter particles to bounce off germanium nuclei in an array of 40, 800-gram cryogenically cooled (18 miliKelvin) ultra-pure germanium crystals, specially built for the experiment at $50k a pop. The crystals are housed inside a "castle" made of ancient Roman lead and radioactively pure polyethylene blocks. The castle is surrounded by a cosmic-ray muon detector, which looks like a giant black box, at high voltage (~2000 V), and the whole contraption sits in a clean-room laboratory under a mountain in Europe. Dr. Evil has a test-bench nearby.
Of course, the best place to do this is under a mountain on the French-Italian border. The five-course dinners and nearby skiing and hiking in Aussois are absolute essentials for a properly conducted low-background terrestrial astroparticle physics experiment. (Seriously, the mountain is necessary.)
EDELWEISS is not the only dark matter experiment that a member of Cloudant has been involved with. Cloudant's Chief Science Officer, Mike Miller, is a member of the team operating the CoGeNT experiment, a direct competitor to EDELWEISS. CoGeNT has produced some spectacular and highly controversial results in the past few years and their spokesperson is somewhat of a lightning rod in the dark matter community. Their results, basically, have been interpreted as signals of dark matter detection!
CoGeNT is not alone. The CRESST and DAMA/LIBRA experiments have also observed signals which have been interpreted to have been caused by dark matter. All three experimental results are controversial because very plausible mundane explanations have been proposed to explain the events they observe, and because of the null results from other experiments. Data from EDELWEISS and others, such as XENON100, XENON10, CDMS, are in direct contradiction because they have not observed any signals that look like dark matter. Yet, the field is so wild and fluid right now even my CDMS colleagues appear to be claiming both! (They're not.)
I should point out that this back-and-forth controversy is exactly what one would expect when on the cusp of discovery! It is certainly a very exciting time in direct dark matter detection and the particle physics community is eagerly awaiting new results. Hopefully new data soon will help us figure this all out.
Before I finish, I want to recognize the EDELWEISS detector fabrication groups in France (Orsay, Saclay and Lyon). The science of low-temperature detectors is an extremely fascinating field and the groups produce absolutely beautiful, state-of-the-art germanium detectors. They are the best in the world and we are very grateful for their work.
Next week I'll tell you about our data and use case. Using Cloudant's services, rather than managing our own systems, helped us #domore physics. And #eatmore raclette.