Rho, Rho, Rho Your Boat

Salutations from the halfway point of our summer research. Week 5 proved to be quite productive, as we really hit the R hard, as well as got into some finer details of atmospheric chemistry with our friend Daniel Jacob and his lovely repository of atmospheric chemical knowledge.

The main objective of this week as far as coding and TDA are concerned was to plot several variables (namely ozone, static air temp, relative humidity, and dew point) against altitude to see if we could observe trends that were expected. Turns out, we did. We also wanted to do some density clustering to see if we could observe features of the atmosphere. To do this, we randomly subsetted the data from the four variables above, used the density cluster tree function of the TDA package, and then replotted the points, but color coded by the corresponding branch in the density tree. This allowed us to clearly see the denser areas of data and then correlate these zones with important parts of the atmosphere, like the ozone layer for example. The plots below show this, with the two main levels of ozone (green and red), corresponding to surface pollution and the ozone layer. These also have some finer detail, like the pink and then gray portion of the top layer showing more precisely where the ozone layer is located.

As far as Jacob's work is concerned, this week was heavy into the finer detail of atmospheric chemical processes. We learned all about the intricacies of the Ox, HOx, and NOx cycles, and how these reactions interact within their own cycles as well as affect the chemistry of each other.

Like I said to begin, this week was very exciting and productive, and both Katie and myself are enjoying getting heavy into the data analysis, and I'm sure it's safe to promise even more exciting and well colored figures are to follow.