The model uses a set of exogenous data loaded at run-time, before simulation starts. This data can be either be endowed with geographical attributes or not.
The formers are GIS layers in ArcInfo ASCII Grid or Grass ASCII Grid format together with an XML meta-data file to describes them. FFSM uses GIS data for administrative borders information (countries and regions), initial forest land cover and forest dynamic modifiers.
Complementary to this data, settings, forest and product data is specified in an Open Document Format for Office Applications (ODF) that can be edited with any modern spreadsheet application.
ODF is a zipped folder of XML files. FFSM++ has the ability to unzip and parse the XML directly, so no spreadsheet installation is required in order to run the model.
The format of this input file is specified in the following table:
| ||List of available scenarios||id, shortDesc, longDesc|
| ||Generic, regional level settings (scalars or vectors)||name, type, comment, value_0, value_1, …|
| ||Overridden settings for the specific scenario||name, type, comment, value_0, value_1, …|
| ||Forest data||parName, region, forType, diamClass, value_0, value_1, …|
| ||Forest data (overridden for the specific scenario)||parName, region, forType, diamClass, value_0, value_1, …|
| ||Products data||parName, region, prod, freeDim, value_0, value_1, …|
| ||Products data (overridden for the specific scenario)||parName, region, prod, freeDim, value_0, value_1, …|
| ||Matrix linking a specific forest type and diameter class to its obtainable wood products||product, forType, dClass, maxYears|
| ||Reclassification rules (see section )||regID, forTypeIn, forTypeOut, coeff|
| ||List of countries and regions under simulation||regId, regSName, regLName, regLevel, parRegId, isResidual|
| ||List forest types under simulation||forTypeId, forLabel, memType, forLayer, ereditatedFrom|
This approach allows to easily manage multiple scenarios keeping a single data source and avoiding data redundancy. In fact, to define a new scenario, it's enough to give it a name it in the scenarios list and change only its specifications in one or more of the
Thanks to the usage of overriding, all the settings and the data in common with the “default” scenario will not need to be repeated. The concept of data overriding is used also in terms of geographical validity of the forest and product data: a variable used in the program at regional level can be introduced in the data at world, national or regional level, depending of the level of information that it's known, but it can be redefined for countries or regions where a more detailed information is available.
For example an elasticity can be expressed at world level, but then redefined at regional level in France and national level in Italy.
Finally, some variables that are expressed as arrays can be interpreted as a time-based series, with the first value being interpreted as the first year.
While this structure allows to keep all scenarios in a single file, it is also possible to keep scenarios in separate files: the location of the input file in fact is selectable either with the -i parameter using FFSM from the command line or with a menu option in the GUI.
FFSM allows for a fair large number of options in its output, the most important is perhaps the output location.
This is specified in the input file in the scenario settings and can be overridden using the
[SCENARIO_NAME]_settings. Therefore each scenario can have its own output directory if this is what is required.
Other variables in the settings sheet allow to tune the output in terms of its verbosity and its layout.
Spatial data can be outputted both as standard images (.PNG) or back as GRASS or ArcInfo ASCII GRID layers to be imported in a GIS for further processing or analysis.
Once printed out in csv format, forest and market data can be handled with a provided ODF spreadsheet that load them and build some pivot tables for an easy analysis (however macro must be enabled on the spreadsheet application for this to work).