Accelerating the modelling of Antarctic snow densification

Polar ice cores from Greenland and Antarctica provide us with a valuable picture on past climate. In order to use the data sets obtained from ice cores is it useful to know something about the transformation of polar surface snow to ice. The Community Firn Model (CFM) is a suite of snow densification models. On top of the CFM we have also built a package that models the diffusion of water isotopes in polar snow (iso-CFM) a process that has can be used as a paleothermometer. Modeling runs especially for the cases of the East Antarctic Plateau can be time-consuming. We want to investigate ways to speed up the model by including parallel processing approaches and/or utilizing GPU support. There is flexibility to place more focus on the software engineering or paleoclimate modeling side of the project depending on the background as well as the wishes of the master student.

Activities: Programming and algorithm analysis

Contact: v.gkinis@nbi.ku.dk

Area: Project Bachelor Masters

Tags: Ice core, GPGPU