Managing Software Complexity and Variability in Coupled Climate Models
Spencer Rugaber, Rocky Dunlap, Leo Mark, Sohrab Ansari · IEEE Software · 2011
Coupled climate models exhibit scientific, numerical, and architectural variability. This variability introduces requirements that give rise to complexity. However, techniques exist that can tame this complexity; one such technique is feature analysis. As climate model fidelity and complexity increase, the climate-modeling community should adopt a systematic way to deal with software variability.