Ideas for optimisation in water use model

[itowers1]

Just noting some ideas to optimise speed of water use model:

• in ff16w strategy, optimum stem water potential/ci is identified, and the profit is recalculated at this value. Instead, could profit be stored iteratively as optimum position is calculated, and then recalled in FF16w strategy
• recalling optimum stem water potential/ci between nodes/timesteps breaks due to required transpiration stream exceeding what is theoretically possible using Newton's method of differentiation, but is likely associated with the fact that the last node of a given time step has very different environmental conditions to the first node of the next time step once the forest has matured a little bit. Thus, storing the recalled value between nodes, rather than across calculations, could potentially solve this issue and unlock Newtons method which is proposed to be faster than GSS.
• calculation of certain values at a high level without having to input them as arguments. For example, the ratio of maximum profits to maximum costs (lambda in sperry's method) is constant for a given node in a given timestep, meaning that it need only be calculated once for a given call to the leaf model. However, it seems unnecessary to have this as another argument akin to k_l_max in calc_profit, thus I am proposing some other way of having this as a pre-computed variable in ff16w strategy which can be passed to leaf model.