%0 Journal Article
%T The Removal of Numerical Drift from Scientific Models
%A John Collins
%A Brian Farrimond
%A David Flower
%A Mark Anderson
%J International Journal of Software Engineering & Applications
%D 2013
%I Academy & Industry Research Collaboration Center (AIRCC)
%X Computer programs often behave differently under different compilers or in different computingenvironments. Relative debugging is a collection of techniques by which these differences are analysed.Differences may arise because ofdifferent interpretations of errors in the code, because of bugs in thecompilers or because of numerical drift, and all of these were observed in the present study. Numericaldrift arises when small and acceptable differences in values computed by different systems areintegrated, so that the results drift apart. This is well understood and need not degrade the validity of theprogram results. Coding errors and compiler bugs may degrade the results and should be removed. Thispaper describes a technique for the comparison of two program runs which removes numerical drift andtherefore exposes coding and compiler errors. The procedure is highly automated and requires very littleintervention by the user. The technique is applied to the Weather Research and Forecasting model, themost widely used weather and climate modelling code
%K Relative debugging
%K numerical drift
%K coding errors
%K compiler errors
%U http://airccse.org/journal/ijsea/papers/4213ijsea04.pdf