Here is a set of R code that fits the growth of grass.

Workflow ideas:

1) Can it be done such that bits of model specification can be done without
	understanding all the R code?

2) Can it be done such that you might think of when to optimally sample the 
	next sample given the current information?
