17 June 2014

Post date: Jun 23, 2014 2:03:51 PM

Emma asked people's opinions on the pattern-oriented modelling ideas from Volker Grimm et al.: see papers in 1996, 2005 (in Science), and 2012.

The general consensus is that it's a good idea (especially when you can calibrate against one pattern and reproduce another 'independently'; they give the Anasazi paper as an example of this in the 2005 Science paper), but that it assumes a particular methodological stance on validation and model purpose (which you need to be sure fits with what you want to do).