Replicating computational atmospheric science

2017-01-03T10:39:40Z (GMT) by James Shaw
<div>Scientific claims can be verified by reproducing or replicating the experiment.  Replication involves new authors writing new code following the original method.  Reproduction involves new authors running the original code.  Doing nothing means that scientific claims remain unverified.</div><div><br></div><div>This talk is based on the work of <a href="https://twitter.com/LorenaABarba">Lorena Barba</a> and was also inspired by <a href="https://twitter.com/philipbstark">Philip Stark</a><a>, </a><a href="https://twitter.com/cboettig">Carl Boettiger</a> and <a href="https://twitter.com/jpeelle">Jonathan Peelle</a>.  I present my own experiences trying to replicate the advection test of <a href="https://doi.org/10.1175/1520-0493(2002)130<2459:ANTFVC>2.0.CO;2">Schär et al. 2002</a>, and the computational reproducibility study of the aerodynamics of flying snakes conducted by <a href="https://arxiv.org/abs/1605.04339">Mesnard and Barba 2016</a>.</div><div><br></div>Presented at the Department of Meteorology PhD group meeting, University of Reading, 9th November 2016.