Data Processing

I must admit that these post titles aren't very interesting, and usually the reason for that is that I am not clever enough to think of something more...well, clever. But this title is specifically intended to convey the mind-numbing process of what they call "modifying."

The potential peaks for one canister (courtesy of Erica Weathers)

Last week, we ran samples using gas chromatography to get signatures for each of the 60+ gases that we can measure. The output of that laborious process is a graph for each canister, with potential difference peaks for each gas. Using a standard, we set the baseline for each gas in each sample and integrated under the curve. This simple sentence does not convey how time-intensive modifying truly is. It made us feel less like hotshot NASA scientists and more like chemistry robots. Which would still be pretty cool, now that I think about it...

Close-up of the peaks for two gases (courtesy of Erica Weathers)

Nick Heath modifying column B
(courtesy of Erica Weathers)

In any case, that occupied the bulk of my week. Yesterday, we matched each sample to the geographical data, like latitude, longitude, altitude and time. That way if we see anything interesting in the data, we know exactly where and when that sample was taken.

Next week, while we all do background research, our PI will compare those areas under the curves obtained from modifying to the standards, and from that calculate the concentration of the gases in every sample. Then comes quality control, where he looks through all of the data and assesses whether any spikes or dips are due to interesting chemistry or bad samples. Once the data is cleaned up, we'll be ready to put together our respective projects to present the following week.