The use of climate simulations in scientific assessments of climate change
and in the formulation of climate change scenarios has been contested for,
among others, methodological reasons. The "philosophy of climate science" e
ncompasses discussions about the methodology of climate science. Three issu
es with respect to climate simulation are discussed: (i) model hierarchy an
d complexity, (ii) tuning and falsifiability, and (iii) uncertainty. In thi
s discussion paper it is argued that high-resolution and low-resolution cli
mate models have complementary roles to play in the science of climate chan
ge. The role of computer simulations in climate science deserves further ph
ilosophical study in order to better assess their quality for informing cli
mate policy making.