Though most people think earth sciences is all about looking at rocks or going on great adventures to remote areas, virtually everyone who works in earth sciences must do computer modeling. This means that most earth scientists have to have knowledge in statistics and even computer science to develop them! I am a geophysicist, meaning that I must delve deeply into the world of programming (perhaps more so that most earth scientists), creating models from basic physical equations and assumptions.
During the past few months, I have been creating ice sheet models using basic equations relating to how ice flows. This required me to create a program written in FORTRAN (one of the oldest programming languages!) that is several thousand lines long! The process of programming not only involves doing the numerical calculation to produce the model, but also to keep track of how the results are stored, and maintaining stability of all the parameters. Programming can sometimes be a painful process, involving many hours tracking down bugs that may not be entirely obvious, and finding ways to make the program run faster. In addition, I found that some techniques that I could use to solve the problem did not work, and I had to investigate alternative ways create the model. Before starting my PHD, I never thought I would have to learn about image processing techniques, like bicubic interpolation!
As you could imagine, there is no greater feeling in the world than when a program you have been working on for months finally works! With the program completed, I can move onto using the models generated from my code to analyze how the earth responds to the loading of ice. The image below shows an example model of the ice thickness of the Laurentide ice sheet from about 20,000 years ago created using my software.