Dynamic Character of FarmDyn
Abstract
FarmDyn can be run in multiple dynamics ranging from the myopic view of a short run/ comparative-static version where only one years is simulated to a fully dynamic version which optimizes over all years. The current version is only tested for the comparative static case
Comparative Static Version and Short-Run
The short-run version considers only one year and does not comprise a liquidation of the enterprise. The comparative static version replaces the herd dynamics by a steady state model where, for example, the cows replaced in the current year are equal to the heifers in the current year, which in turn are equal to the calves raised in the current year. In the comparative static mode, the vintage model for investments in buildings and machinery is replaced by a setting in which the investment costs are related to one year. Nevertheless, the binary character can be maintained.
The Fully Dynamic Version
The fully dynamic version optimises the farm production process over time in a fully dynamic setting, i.e. all time points are simultaneously considered. Connecting different modules over time (t1-tn) allows considering biologic and economic path dependencies.
The temporal resolution varies across different parts of the template module. Cropping decisions are annually implemented, whereas the intra-year resolution of the herd size module can be flexibly chosen by the user with a minimal resolution of one month.
Concerning fodder composition, decision points in each year are every three months. This provides the decision maker a more flexible adjustment to feed requirements of the herd (conditional on lactation phase), his resources and prices respectively availability of pasture, silage and concentrates. Furthermore, as stated in the manure module, the applications of manure or synthetic fertilisers, as well as the stored manure amounts on farm are implemented on monthly level.
The optimal production plan over time is not simulated in a recursive fashion from year to year, but all variables of the planning horizon are optimised at once. Consequently, decisions at some point in time also influence decisions before and not only after that point. For instance, an increase in the herd at some point might require increased raising processes before.