One of the major research goals of the lab are to understand the nature and interactions of cloud particles. Most of my work concerned experiments that recreate cloud conditions in the laboratory using a turbulence chamber. In it, water droplets (~100 μm in diameter) were subjected to various levels of turbulence and observed using inline holography with two high speed cameras. This is done so that the particles can be tracked and useful statistics can be found such as fall speeds, particle collision rates, etc. These will hopefully be able to increase our understanding of clouds. An example of a hologram taken by one of the cameras and a reconstruction of it at the middle of the chamber are shown here. Notice the particles in focus or close to in focus in the reconstruction.

These experiments using the turbulence chamber involved optics/holography, camera calibration, particle matching, particle tracking, particle statistics, and miscellaneous tasks. I worked on camera calibration, particle matching, a little particle tracking, and some of the many miscellaneous tasks.

My most important accomplishments were successful camera calibration and decent particle matching. The camera calibration involved calculating the positions of the two cameras relative to each other. Rather than setting up a calibration run with objects at known positions to calibrate the cameras, the actual data from the runs were used to calculate the camera positions. This step lead to the next step: particle matching. This involved matching the particles between the two cameras and calculating a best estimate of their positions using information from both cameras. I basically took the positions of the particles seen be each camera individually, calculate the cameras' positions, and found the particles seen by both cameras. This is important as the depth uncertainty in holograms is quite large and two cameras can be used to improve the resolution, which helps tracking and the various statistics greatly. Two plots of some tracks are shown here. One of them shows the tracks obtained just using the information from 1 camera. The other uses the information from both cameras (particles matched between them).

The miscellaneous tasks were very varied. A few of the things included mini experiments to figure out what property of our laser was unstable, figuring out how to use another set of cameras to verify what was seen in the holograms, figuring out the cameras didn't start at the same time, finding out that the laser beam was diverging, etc. This lead to many relevant pieces of information such as the holography laser's beam was diverging and many non-related but interesting pieces of information. For example, a graduate student in the lab found a really clever way to roughly measure the speed of electricity in a wire (it came out to be about 2/3 the speed of light). In short, research often involves more time working on the apparatus, finding and fixing problems, double checking, learning sometimes unrelated things, etc. than collecting and analyzing data. If it wasn't this way, research wouldn't be as interesting.