European Forestry Dynamics Model
The European Forestry Dynamics Model was developed in R . In its current version, it is designed to model even-aged forests (a version of EFDM to deal with non even-aged forests is currently under development, with a foreseen completion in the fall of 2015). Using a specification of management as input, EFDM simulates the development of the forest and estimates volume of wood harvested. This estimate can be broken down per category modelled, for example species, site quality, management regime etc. EFDM was jointly developed by Metla  (now LUKE ) and the Swedish University of Agricultural Sciences  for the Forest Resources and Climate Unit of the European Commission's Joint Research Centre  in the framework of the Specific Contract 10 under FWC388432 in 2013, and was further enhanced under Specific Contract 14 of the same FWC under the expertise of the same developing team. The model is currently functional.
EFDM was conceived as a flexible system for harmonized forestry modelling for all European countries. It is intended to process data from National Forest Inventories. This data is not standardized, nor is it necessarily available outside of its parent country, hence the need for a modular system such as EFDM.
EFDM is in its early stages. With the help of a growing community of users, we hope it will become a full-fledged software tool in the near future. In addition to software taking care of the core functionally of the model, there are components calculating the final felled or thinned volume of wood over a given time period (see Output page); it also contains of a series of functions, helping user set up their input data.
The EFDM package is now available as free software licensed under the *European* Union Public Licence (EUPL) v1.1. EUPLv1.1 is compatible with several free software licenses.
The EFDM is built around a basic matrix structure, defined by a set of fixed states, between which “units” of forest move over time. The core of the matrix structure is the dynamic state-space, which is defined by two variables: volume and age. Different dynamic state-spaces can be set up for different forest types. which could be defined by the modeller using variables as site quality, geographical region, species (group), owner type etc. All dynamics within the model take place inside the different “dynamic state-spaces” , thus no shifts between the forestry types are currently modelled. However an area could theoretically move from one forestry type to another, for example, as a result of forest regeneration with a different tree species.
At every simulation step, the distribution of the forest area in the state-space changes. The change is driven by the transition probabilities. Transitions are made within one time step in EFDM in the following sequence: first, the areas in each cell of the state-space are split according to activity probabilities. In the next step, the distributions of the areas in the activity-specific state-space are modified according to transition probabilities for that activity. In the last step, the activity-specific state-spaces are merged back into one, and their growth progresses without any "memory" of the activities that have been applied. The resulting distribution of forest area in the state-space describes the state of the forest after one time step. In the next cycle, the model then applies the the activities probabilities on the "new" state-space.
How to install EFDM
- Download v.2 of the code from our main repository here:  (The v.1 is also available here, as well as from a third-party site here:).
- Install R  with "abind" package (see Installing abind package for R); and R-Studio 
- Open R-Studio
- Make a new project by clicking on upper right icon "New Project"
- Choose a new directory
- Choose an empty project
- Give a name to your project and choose the directory
- Put the contents from the "EFDMcode" folder on the repository in this directory.
Getting started with a toy dataset
You may also transfer a toy dataset from the "toyDataSet" folders of the EFDM repository to the folder you created in the previous steps. These datasets were built to test the EFDM and are not necessarily representative of the current situation in their countries' respective forests and their evolution. In the folders there are several files that can be transferred to your R project folder (in the same folder as the EFDM model code). The key files are:
- where probability of thinning is estimated: thinP.RData
1. In the RStudio prompt, type: >source("efdmutils.r") >runefdm("efdminput.txt") OR >efdmui() *If you choose to run EFDM with the input text file using the 'runefdm' command,the results will be written to file without any need for further interaction. **Remember to change the path names in input files** *If you choose to run EFDM interactively using the "efdmui" command, the following steps should be followed: 2. You will be prompted to choose a file that includes Factors -> choose "factors.txt" from the toy dataset 3. You will be prompted to choose a file that corresponds to the forest Initial State -> choose "initstate.txt" from the toy dataset 4. You will be prompted to choose a file that summarises the Activities in your forest (both management and natural processes) -> choose "activities.txt" from the toy dataset 5. You will be prompted to choose a file that corresponds to the Activities' Probabilities that each of these activities occurs -> choose "actprobs.txt" from the toy dataset 6. You will be prompted to choose how many Times Steps you wish to run the model -> choose any integer, keeping in mind that in this toy dataset,each time step corresponds to 5 years. 7. EFDM will then ask you if you want output in addition to raw results. The default output is a large table (see "rawoutput.txt") with a description of the forest area per activity type per time step per forest class. If the user would like the output data to be fancier, they can type a "1" at the prompt (1=yes), however this implies that an output request file is prepared in advance. This is the case for the toy dataset ("outputrequests.txt"). 8. Based on the above steps, EFDM has written a summary of instructions to a file called "efdminput.txt". This file can be manually edited to change any of the choices made previously. 9. The model is launched with the following command: >runefdm("efdminput.txt") 10. Your output is written to the same location as your input. -> the main table of raw output is called "rawoutput.txt" -> if you requested any plots, they are jpegs
Setting up your own data
- The first step in setting up a dataset for EFDM is to choose the most appropriate model set up. This is analogous to choosing a set of factors. This concept is discussed in depth here: Factors.
- The forest's initial state must then be described. Details on how exactly to do this can be found here: Initial State.
- The activities you wish to simulate must be described. As a start, you may want to model a "no management" scenario to get a feel of how the model works. See Activities to set up EFDM to include management practices and/or natural phenomena such a forest fires or pests.
- The activities must be associated with a probability of occurrence. This is described in the Activities' Probabilities matrix.
- Transition matrices for different activities should be constructed. There are several ways to do this, these are described in the Transition Probabilities page.
- A request for a specific format of output can be made, for example user-specified graphs can be produced. More detail on this is given in the output request file page.
PACKALEN Tuula; SALLNAES Ola; SIRKIA Seija; KORHONEN Kari; SALMINEN Olli; VIDAL CLAUDE; ROBERT Nicolas; COLIN Antoine; BELOUARD Thierry; SCHADAUER Klemens; BERGER Ambros; REGO Francisco; LOURO Graca; CAMIA Andrea; RÄTY MINNA; SAN-MIGUEL-AYANZ Jesus, 2014. "The European Forestry Dynamics Model: Concept, design and results of first case studies". Publications Office of the European Union, EUR 27004 doi 10.2788/153990 
The context of EFDM tool development and some description of the model can be found in Mubareka,S., Jonsson, R., Rinaldi, F., Fiorese, G., San-Miguel-Ayanz, J., Sallnas, O., Baranzelli, C., Pilli, R., Lavalle, C., Kitous, A. 2014 An Integrated Modelling Framework for the Forest-based Bioeconomy. 
Reference material in printable format is available on the software repository under "documents" .