AMS's science goals are things like "search for anti-helium", "measure positron spectrum", "search for strangelets". But the data it transmits to the ground looks more like this:
| Tracker | strip #12334 | 123 units |
| Tracker | strip #12354 | 14 units |
| Tracker | strip #18003 | 1024 units |
| Tracker | strip #00334 | 54 units |
| TOF | tube #54 | 7 units |
| TRD | temp #100 | 10 C |
Understanding the data begins at the Science Operations Center, or S.O.C., where we pore over the raw data. We have to combine all of our knowledge of physics, our years of detector construction and calibration, vast computer simulationsm and the properties of the flight data itself, to understand the physics. Ultimately, we end up with a suite of software - called "reconstruction software" - which processes raw data (like the made-up example above), and generates a database of particles. Each database entry lists one particle that passed through the detector; it contains our best guesses as to that particle's charge, speed, momentum, etc.; and it contains a lot of detailed detector information.
The reconstructed particle database contains a mixture of well-measured particle events, and poorly-measured events where (for example) there are large error bars on the mass, or the charge, and even events where we're not sure we saw a particle at all. At the Science Operations Center, AMS physicists will also wrestle with these uncertainties. We will find the best way to measure the physical processes - the antihelium nuclei, cosmic-ray positrons or whatever - that are responsible for every one of those bits of raw data.