Gravner and Griffeath have developed a serial numerical model that simulates the growth of snowflakes in three dimensions under the assumptions of 24-fold symmetry. To allow for much larger and asymmetric snowflakes as well as to reduce simulation time, their model was re-implemented using distributed parallelism via MPI (Message Passing Interface). Test-driven development (TDD) was applied to rapidly develop an accurate implementation that consistently reproduces the results of Gravner and Griffeath. Through parallelism, simulation times were reduced from days/weeks to mere hours, and crystal sizes could be explored that are roughly 10 times larger in each dimension than otherwise possible. This new implementation will be used to generate thousands of representative snow crystals as a means to improve the ability to use remote sensing to estimate water content in snow-containing clouds.

This work was done by Thomas Clune of Goddard Space Flight Center, and Christopher Pearson of Northrup Grumman Information Technology. GSC-16346-1

NASA Tech Briefs Magazine

This article first appeared in the March, 2016 issue of NASA Tech Briefs Magazine.

Read more articles from this issue here.

Read more articles from the archives here.