||Climate Change Analysis
From the REU Project Description page:
I am interested in refining the climate change analyses described in a "weather module" that is being developed for the Mathematics of Planet Earth 2013 project. In particular, it would be interesting to make a simplified version of the world climate change map in which each location is analyzed not by solving a daily model with sinusoidal seasonal fluctuations but rather a simpler model that uses annual averages for the input to a linear regression model. I will be working with Dr. Robert Vanderbei this summer. Dr. Vanderbei is a leading researcher in operations research and financial engineering. He is also an associated faculty member with the Astrophysics, Computer Science, Mathematics, and Mechanical and Aerospace Engineering departments at Princeton Universtiy
- Week 1:
- So here I am at Rutgers working on my first ever research project! I spent the first half of the week just exploring the campus and its surroundings as I was yet to meet my mentor. On Thursday, I went to Princeton University to meet my mentor, Dr. Robert Vanderbei, and discussed some possibilities for the summer. Then I worked on my presentation which went well on Friday.
- Week 2:
- I went through some of my mentor's past works on the topic. He hooked me up with an access to AMPL, A Mathematical Programming Language (literally!). He also sent me most of the required data files so that I didn't have to dig into the NOAA database (Thank you Dr. Vanderbei!). I spent rest of the week getting familiar with the language and attending cool seminars.
- Week 3:
- I did the first round of data analysis. Using NCAR Command Language, I plotted the weather-station locations scattered throughout the world in a spherical globe. Then I created Delaunay triangulation and Voronoi polygons from them. (A voronoi Polygon is a polygon whose interior consists of all points in the plane which are closer to a particular point than to any other.) One of my goals was to compute the average change in temperature on the surface of earth per century. For that, I used the change in each station weighted by the area they represent. However, the computed average was not really close to what was expected. Time to rethink the process!
- Week 4:
- More seminars! And an exciting cultural day! It was a fun week and between all those, I managed to get to Princeton once again and talked to my mentor regarding the not-so-accurate results. We figured out the possible causes and planned out a better method of computation. Then my mentor mentioned something interesting. While we are here, spending entire summer to calculate how much the temperature has risen due to global warming, a New York Times article says that there has been no "warming" in last 20 years! Hmm.. point to be noted. We decided to give it a try and see if we get the same answer from the data we have for last 20 years.
- Week 5:
- Thanks to Wimbledon finals and Fourth of July, I was not able to do much work this week. (And yay, Murray won!!) I managed to color the polygons on the globe I mentioned earlier with a range of colors representing the change of temperature in each station. I also separated out the data for last 20 years from the gigantic data files I had; now I just need to feed them to a shell script and the results will follow very soon.
- Week 6:
- The results turned out to be consistent with the article. Global warming has indeed been passive for last 15-20 years (Praise the almighty!). Another surprising result that I got is that there had been a global cooling instead of global warming during the mid 20th century. As it turns out, that period is referred to as a little ice age! These "cool" results have slightly changed the plans for rest of the program.