iForest - Biogeosciences and Forestry

Integrating forest-based industry and forest resource modeling

iForest - Biogeosciences and Forestry, Volume 9, Issue 5, Pages 743-750 (2016)
doi: https://doi.org/10.3832/ifor1961-009
Published: Aug 12, 2016 - Copyright © 2016 SISEF

Research Articles

This paper introduces a modeling approach for the assessment of policy options within the forest-based bioeconomy. The feedback between the forestry dynamics model and the economic model of the global forest-based sector of the proposed framework is essential, not only for response analysis as to the development of forest resources and for a correct assessment of future harvesting potentials, but also for the assessment of the impact of different management regimes on wood-based product markets. Test runs of the modeling framework on a Swedish case highlight the necessity of considering timber assortments for a comprehensive integration of forest resources and wood-based commodity market dynamics. Hence, the composition of harvest demand in terms of timber assortment affects the allocation of harvesting activities and, consequently, the development of forest resources (and thus future harvest potentials), as well as the production, trade and consumption of wood-based products.

Policy, Bioeconomy, Wood-based Products, Market, Forest Resources


A meaningful assessment of the impact of policy actions within the forest-based bioeconomy requires the capability to evaluate the economic implications on the market for wood-based products, as well as the impact of such implications on forest resources. This in turn calls for the modeling of the wood-based product market, as well as of the forest resource dynamics, and, most importantly, their interaction. While there are quite a few modeling efforts with pan-European scope that dealt with forest resource dynamics ([13], [23], [24], [3]), or wood product markets separately ([20], [21], [9], [12]), only few modeled their interaction on pan-European scale ([10], [19], [22]).

In particular, the existing studies do not fully account for the feedback from the forest-products market model to the forest resource model, so that the integration between the two is partial only. Thus, the sustainable potential supply of woody biomass as calculated by the forest resource model is ingested by the market model as a constraint on the production of wood-based products (sawnwood, wood-based panels, pulp and paper). The demand for wood raw material calculated by the latter is then used by the forest resource model to assess the development of forest resources; however, when computing the next period potential harvest level, the “actual” harvest demand derived by the market model is not taken into account.

As a result, multiple errors propagate over time, should the satisfaction of the demand for woody biomass calculated by the economic forest sector model require a lower harvesting level than the sustainable potential derived by the forest resource model ([15]). In addition, none of these studies accounted for timber assortments (sawlogs and pulp-/fuel-wood, respectively), neither when allocating the harvest demand from the market model in the forest model, nor when deriving in the latter the harvest potential to be used as bounds for the production.

On the opposite, the full integration of forest resources and market dynamics in a modeling framework calls for the consideration of timber assortments. Hence, the division of harvest potential on timber assortments has implication for production as well as for trade of wood-based products. In turn, the composition of the harvest demand in terms of timber assortment has implications for forest management and, consequently, for the development of forest resources.

This paper adds to the existing literature by elaborating the full (as opposed to the partial one) interaction between a forest resource model and an economic forest-based sector model. Further, in the information transfer between the two models, timber assortments (coniferous and non-coniferous sawlogs and pulp-/fuel-wood, respectively) are also accounted for. The integrated model is used in a Swedish test case.

The paper proceeds as follows: the next chapter introduces the forest-based sector model - the Global Forest Trade Model (GFTM - [7]) - and the forest resource model - the European Forestry Dynamics Model (EFDM - [14]). Then a description of the information exchange between the two models follows. Hereafter, the results of a number of modeling runs are presented and discussed. Finally, conclusions and suggestions for further research are put forward.

  Materials and methods 

The modeling framework for the forest-based bioeconomy suggested in this paper is based on two main ingredients: a forest-based sector model - the Global Forest Trade Model (GFTM) - and a forest resource model - the European Forestry Dynamics Model (EFDM).

The Global Forest Trade Model (GFTM)

The Global Forest Trade Model (GFTM) is an equilibrium model for the forest-based sector which shares with other similar models - notably the Global Forest Products Model, GFPM ([2]) and the European Forest Institute Global Trade Model, EFI-GTM ([8]) - the theoretical formulation based on spatial equilibrium theory in competitive markets for several commodities ([18]). Specifically, the model is based on the maximization of the whole forest sector welfare (consumer, primary/industrial products-producers and traders), subject to feasibility, resources, productivity and equilibrium constraints. Similarly to the GFPM and the EFI-GTM, also the GFTM is static since, given a certain number of iterations (i.e., the number of periods one wants to project), at each iteration the optimal welfare is computed, with imperfect foresight. Once a solution is reached, the parameters of the model are updated based on endogenous (harvest levels) and exogenous (GDP growth) drivers, new resources and productivity constraints are set, and a new iteration begins.

GFTM focuses on wood-based products that are internationally traded, covering ten final products, four intermediate products, and four primary products (Fig. 1). The geographical scope is global, with a European focus. Countries that are modeled individually comprise all EU member states plus Belarus, Norway, the Russian Federation, Serbia, Switzerland, and Ukraine in Europe, and then Brazil, Canada, Chile, China, India, Japan, Turkey, and USA. Some non-major producer and/or consumer countries of wood are aggregated into global sub-regions: South East Asia, North Africa, South Africa, Rest of Latin America, Oceania, and the Rest of the World. The main outputs of GFTM are the projections of consumption, production and net trade levels for final products, the projections for harvested, industrially processed and net traded quantities for primary products, and quantities produced and traded for intermediate products.

Fig. 1 - Product flow chart of the GFTM.

  Enlarge/Shrink   Download   Full Width  Open in Viewer

GFTM uses as starting prices/values for production/trade quantities the corresponding data derived from FAOSTAT and EUROSTAT data bases for the years 2010 and 2011. Prices for all commodities derive from trade unit values (value in US$ divided by quantity exported or imported). Following the same approach as for the time series cross sectional approach in Jonsson ([6]), the largest trade stream (in quantity terms) is used to derive the price (e.g., for the Swedish price of coniferous sawnwood, the export trade unit value is used as the price for sawnwood). Production costs derive from FAOSTAT as the price (trade unit values) for the product minus the price of the input(s) weighted by input coefficients.

Price and GDP elasticities derive from Jonsson ([6]) for Europe, while for non-European countries and sub-regions GFTM uses the same elasticities as in GFPM. As for wood pellets, price and GDP elasticities are estimated through time-series cross-sectional analysis performed on data for household use in Austria, Germany, Italy, and Sweden. These elasticities are then applied for countries where the consumption of wood pellets is deemed to be dominated by household use, while for countries where wood pellets are consumed also for larger scale use for heating and/or power, weighted elasticities are set based on expert assessment of the respective quantity share of respective user category. The constant parameters of the timber supply function in the initial period derive from actual data for sawlogs and pulpwood removals (production), and prices of sawlogs and pulpwood, respectively.

GFTM can be used as a stand-alone model. In this instance, timber supply is provided from a simple growth model wherein growing stock and increment data are compiled from various sources, and annual potential harvest levels are set equal to annual increment. This volume is then converted to solid volume under bark, using a constant 0.88. This value is based on national data for Sweden, considering all tree species and assortments. Ideally one should use different conversion factors depending on species. In particular, non-coniferous species are very diverse in terms of the extent of bark. However, for simplicity, considering that all non-coniferous species are aggregated in the modeling, only one conversion factor was used here. Finally, the obtained volume is divided into coniferous and non-coniferous, sawlogs and pulpwood respectively, based on FAOSTAT industrial roundwood production (i.e., removals) data series.

The supply/availability of local intermediate and final products is determined in the transformation process simulated by the industry module of the GFTM. Thus, the transformation of products implicit in the production process is described in GFTM by means of a country-specific industry matrix, whose number of columns equalizes the number of produced products (intermediate and final), the number of rows is the total number of products, and the matrix coefficients are equal to the conversion factors for production. Unfortunately, for most countries and products these data are not available, hence the used input/output coefficients build on Fonseca ([5]), while, for countries and sub-regions not covered in the study, the coefficients were extrapolated using expert assessment.

For further details regarding model structure, assumptions, and input data, we refer the reader to Jonsson et al. ([7]).

European Forestry Dynamics Model (EFDM)

EFDM is an area-based matrix model ([16]), meaning that forest areas are transiting between elements of a set of fixed states, depending on the initial state and the correspondingly applied management activities. Typically, an activity is either harvest, such as final felling and thinning, or no management, which simply means that the forest is let to grow naturally for that time step. However, an activity could also represent a calamity due to biotic or abiotic forest damage. Thus, given a set of fixed states S, and denoting by X0 the initial area distribution over the states, by P the transitions between different states (S) guided by the activities A (defined over S), and by t the ordinal number of time step in the model run, the transition from one period to the following are governed by (eqn. 1):

\begin{equation} X_{t+1} =\left(\sum_{j}P_{j}A_{j}\right)X_{t} \end{equation}

When applied to even-aged forests, the set S is usually defined by classes for age and standing volume. A common S is associated with all the different “forest types” which in turn are defined by, for example, region, species, site quality and/or owner. The initial state matrix X0 is estimated using NFI plot data, while the transition matrix P is estimated using two consecutive measurements of NFI plots, increment measurement of NFI plots, or growth information from pre-existing functions.

Usually, data available for estimating the transition matrices also define the length of the time step of the model. The X matrix is constructed through a simple classification, while the P matrix is estimated using a Bayesian procedure. The activity matrix A is derived relying on national expertise and it consists of probabilities for each activity in each cell of the state-space, normally expressing a management pattern relating to the “handbook”. By applying a shifter to the basic activity matrix, the activity level can be adjusted to meet, for example, a specific harvest level. The relative intensities of different activities can be changed using activity specific shifters.

In the Swedish test case used in this paper to demonstrate the full feedback loop, the set S was defined by 10 (11) volume-classes and 32 (33) age-classes for 36 different “forest types” (four site classes, three species and three owner groups) in all four different regions covering entire country. Volume and age classes are considered to be dynamic, while the factors defining the forest types are static. For details about the forest type definitions see Tab. 1.

Tab. 1 - The static factors defining the forest types in Sweden.

Owner State Private Company -
Dominating species Pine Spruce Broadleaves -
Region Götaland Svealand S. Norrland N. Norrland
Site class
(m3 ha-1 year-1)
1 < 7 < 4.4 < 2.9 < 2.5
2 7.0 - 8.5 4.5 -5.7 3.0 - 3.5 2.6 - 3.1
3 8.6 - 10.0 5.8 - 7.1 3.6 - 4.2 3.2 - 3.7
4 ≥ 10.1 ≥ 7.2 ≥ 4.3 ≥ 3.8

  Enlarge/Reduce  Open in Viewer

Model linkages

In each period the iteration begins with the forest resource model, EFDM, providing the maximal sustainable harvesting level in terms of coniferous and non-coniferous sawlogs and pulpwood to GFTM. However, for countries not (yet) modeled by EFDM, the simpler model mentioned above is used to provide the maximal sustainable harvesting level. Although EFDM provides the supply of timber split in sawlogs and pulpwood, the real proportion is derived in the GFTM depending on the demand for wood-based products and wood pellets ([7]).

Thus, the factual distribution on different assortments of a certain harvested amount is dependent on market conditions (i.e., depending on demand, some sawlogs can be used as pulpwood), and hence derived within the GFTM. However, a possible assortment outcome of the harvest is critical in terms of impact analysis, since it allows identifying the implications for the market of a change in management practices resulting from policy interventions, e.g., a shift from a clear felling regime to continuous cover forest management.

GFTM reaches an equilibrium solution, and it provides, besides other results, the demand for coniferous and non-coniferous sawlogs and pulpwood specified for each country (or global sub-region). The EFDM ingests such demand, and allocates the harvesting activities in the country (or sub-region) to meet the demand for the respective assortment. The new state of the forest resulting from the harvest and other management activities is then calculated, and a new harvest potential is provided, while demand curves are updated using growth projections for gross domestic product (GDP).

Basic outputs of EFDM are the distribution of forest land area into volume and age classes after a time step and the drain per activity during the time step. However, in this application the drains were converted into timber assortments. Thus, for each forest type in the forest resource model, there is an assortment table that provides the outcome in terms of assortments for each state-space cell and activity.

An output (assortment) vector was established for each forest type and activity. Linking these vectors to the activity shifters, the allocation of harvest demand was solved by finding the combination of sub-region, species and activity specific shifters that would yield an outcome close to the demand. The same tables were used to express the potential harvest level, which was established in terms of plain cubic meters, in assortment volumes.

Establishing such an assortment table for a given country requires primarily some information about the theoretical, potential, distribution of assortments in the given forest type volume-age-cell. This theoretical distribution is contingent upon local conditions, besides bio-physical ditto, mainly minimum top-diameter requirements for sawlogs and pulpwood. Thus, the theoretical distribution is then affected by harvesting methods, equipment, and traditions.

Thus, the creation of - as realistic as possible - assortment tables presupposes the involvement of national expertise. In the Swedish test case, these tables were derived from the output of a national forestry planning and analysis system Heureka ([11]), which implies that the tables provide the theoretically possible assortment distribution, rather than the market driven distribution. Hence, we were not able to estimate assortments actually entering the market.

Further, since the forest type definitions are based on dominant species, a fraction of the wood harvested in a coniferous forest type could actually originate from a non-conifer species, and vice versa. This was handled through coefficients expressing the fraction of the volume harvested in a forest type in a certain activity that would be allocated to non-conifer assortments.

The potential harvest level is derived inside the forest resource model as the highest possible (harvest) level that could be sustained for 100 years without significantly decreasing the standing volume. In the test case, the harvest level was allowed to be temporarily 10% lower than the long-term level, and a standing volume 10% lower than the initial one was also accepted. In these calculations, a standard management pattern was assumed and a common shifter was used for all sub-regions, species and activities. In cases where the total sum of all activities for a cell in a state-space would have exceeded 100%, the proportions were adjusted favoring final felling, then thinning, and lastly no management.


Two different sets of model runs were performed. In the first one, the flow between the GFTM and EFDM was specified in assortments, while in the second the split of the harvest potential for Sweden provided by EFDM was done based on historical industrial roundwood removals data. In both cases the GFTM was run on a global scale (48 regions). The simpler growth model mentioned above was used as forest component, except for Sweden for which EFDM was employed. Specifically, the two forest models have been used to provide, for Sweden and for 47 regions, respectively, the harvest potential, and then, after GFTM’s run, to ingest the harvested demand when computing the following period’s harvest potential.

Tab. 2 depicts the maximal sustainable harvesting level - split in coniferous and non-coniferous sawlogs and pulpwood, respectively - derived by the EFDM and ingested by GFTM as upper bounds for the production of wood-based commodities, the resulting demand for sawlogs and pulpwood provided by GFTM as a feedback to EFDM, the harvest potential for the next period, and the ensuing demand for primary products. As the sustainable harvest potentials provided - the initial one as well as the following ones revised in each (five-year) period - are higher than the demand for all assortments, the potentials, in total as well as per assortment, are increasing. There is further a noticeable reduction in the modeled demand for coniferous sawlogs (Csl) over time. The demand for coniferous pulpwood (Cpw) is also slightly reduced, whereas the demand for non-coniferous sawlogs (NCsl) and non-coniferous pulpwood (NCpw) increases.

Tab. 2 - Harvest potential and harvest levels in Sweden as modelled by EFDM and GFTM respectively when assortments are accounted for in EFDM.

Annual harvest potential (million m3) - EFDM Annual harvest levels (million m3) - GFTM
Period Csl Cpw NCsl NCpw Sum Period Csl Cpw NCsl NCpw Sum
2010 43.1 23.9 1.1 10.1 78.2 2010 36.8 23.4 0.36 3.3 64.0
2011-2015 44.2 24.9 1.3 11.4 81.8 2011-2015 36.3 23.2 0.37 3.5 63.3
2016-2020 45.5 26.0 1.4 13.0 85.9 2016-2020 35.4 23.2 0.38 3.7 62.6
2021-2025 47.4 27.5 1.6 14.1 90.6 2021-2025 34.3 23.4 0.40 3.8 62.0

  Enlarge/Reduce  Open in Viewer

Tab. 3 is the correspondent to Tab. 2 for the case where timber assortments are not accounted for, neither when allocating harvesting activities, nor when providing the harvest potential, to GFTM (which in this instance is provided as total industrial roundwood). Harvest potentials, as well as harvest levels, differ from the corresponding ones derived when timber assortments are accounted for in the EFDM. Interestingly, the total harvest potentials obtained after 2010 are higher when the EFDM does not consider timber assortments (Tab. 3), even though total harvest, obtained summing up over all assortments and periods, is higher than the one derived when timber assortments are taken into account.

Tab. 3 - Harvest potential and harvest levels in Sweden as modelled by EFDM and GFTM respectively when assortments are not accounted for in EFDM.

Annual harvest potential (million m3) - EFDM Annual harvest levels (million m3) - GFTM
Period Conifers Non-Conifers Sum Period Csl Cpw NCsl NCpw Sum
2010 67.0 11.2 78.2 2010 33.3 28.6 0.36 4.0 66.2
2011-2015 70.6 12.5 83.2 2011-2015 32.1 29.3 0.37 4.3 66.1
2016-2020 72.2 14.2 86.5 2016-2020 32.0 25.3 0.38 3.8 61.4
2021-2025 78.2 15.6 93.8 2021-2025 30.8 26.0 0.38 4.0 61.2

  Enlarge/Reduce  Open in Viewer

As regards Sweden, Tab. 4 largely mirrors the information in Tab. 2. Hence, most notable is the slight, gradual reduction in coniferous sawnwood (Csw) production in Sweden, whereas the production of non-coniferous sawnwood (NCsw), plywood (Pw), particle board (Pb), fibreboard (Fb), graphical paper (GP), packaging paper (PP), and household & sanitary paper (HH&SP) remains largely unchanged.

Tab. 4 - Annual production levels of wood-based commodities as modeled by GFTM (m3/tons) when assortments are accounted for in EFDM.

Year Assortments SWE Fin Fra Ger UK Russia Canada USA China
2010 Csw 17891166 9990261 7650173 21296513 3411216 29808323 38580341 38897585 18690340
NCsw 136268 77914 1535976 1041425 55645 2557317 1533942 17697594 24338855
Pw 73469 954475 270961 232503 0 3035975 1795289 9345798 45544028
Pb 491376 209131 4198261 7295367 2486826 5483644 6243751 15072684 11999835
Fb 96849 99961 1070536 4621156 706642 1700114 1274935 7601422 42658061
GP 5169501 7252401 3573665 10171953 1627377 2456551 8575889 20477454 26900999
PP 4997007 4380371 4426774 11571165 1824451 2859312 3333179 53547196 64693252
HH&SP 338698 151026 701449 1312430 753204 308350 621155 6577437 7871030
2015 Csw 17613448 10454891 7946522 21940973 3516001 30608135 39661070 40128492 19320655
NCsw 139364 79269 1578847 1074507 58539 2613065 1582117 18432866 25060511
Pw 73357 938591 269989 232093 0 3050455 1795732 9327121 46270246
Pb 491338 208637 4218638 7279527 2446079 5418181 6274890 14951124 12524843
Fb 96812 99946 1071906 4680061 681143 1691143 1270790 7579650 43972106
GP 5138924 7117461 3556903 10169209 1612282 2482819 8606936 20095269 27403120
PP 4998955 4605787 4390459 11602165 1749151 2811079 3329248 56214207 67254192
HH&SP 328416 147203 671764 1268612 742095 307144 539614 6371646 8216248
2020 Csw 17146123 10804002 8192772 22508058 3627119 30972095 40513646 40859709 19726795
NCsw 143975 81664 1625453 1104500 62575 2675076 1627838 18965205 25446259
Pw 73353 939525 270225 232322 0 3049267 1798885 9314515 46883691
Pb 491009 208503 4232874 7286865 2403710 5370395 6293746 14744510 13067557
Fb 96814 99946 1073666 4734449 618512 1686427 1269119 7539897 44922030
GP 5105685 7038596 3546049 10167224 1591936 2489075 8631845 19834590 27977269
PP 5025373 4779517 4375406 11619047 1715743 2789099 3322526 58975108 69315619
HH&SP 325885 146453 666494 1254271 737875 306784 475898 6295120 8647999
2025 Csw 16615814 11169239 8407304 22963675 3711211 31843158 42105484 42459647 19942088
NCsw 147647 83635 1664538 1132685 65427 2734833 1668134 19395230 25866740
Pw 73355 941731 270574 232699 0 3061702 1806218 9306326 47754089
Pb 490890 208364 4246859 7285879 2375692 5337334 6312012 14775913 13635293
Fb 96830 99954 1076703 4781277 605397 1683664 1268439 7594958 45976764
GP 5101626 6964843 3549313 10236367 1579283 2502254 8674228 19724387 28832250
PP 5112092 4974170 4370206 11733887 1690495 2771663 3337275 61884964 71632311
HH&SP 323360 145699 660573 1241347 733691 306458 430555 6332634 9180277

  Enlarge/Reduce  Open in Viewer

Tab. 5 is the correspondent to Tab. 4 for the case where timber assortments are not included in the EFDM. There are notable differences in the production figures, not only for Sweden. The most pertinent ones are the lower production levels of coniferous sawnwood in Sweden as compared to the case with assortments, and the correspondingly higher production in some other major producer countries, notably Canada, Germany, the Russian Federation and the USA. The lower production levels of coniferous sawnwood in Sweden can be understood in light of the historically derived coniferous sawlog proportion (53% of industrial conifer roundwood removals) used to split the harvest potential in assortments in these runs, being markedly lower than the one provided by EFDM in the runs considering timber assortments. In turn, this difference is an effect of the method used for creating the assortment tables in EFDM, as discussed in the Materials and Methods.

Tab. 5 - Annual production levels of wood-based commodities as modelled by GFTM (m3/tons) when assortments are not accounted for in EFDM.

Year Assortments SWE Fin Fra Ger UK Russia Canada USA China
2010 Csw 16101652 9998495 7646496 21287558 3410978 30055339 38910992 39089895 18782111
NCsw 135536 77876 1536404 1042435 55613 2557375 1534490 17695829 24458212
Pw 73335 951117 270940 232595 0 3035771 1793802 9341244 45492572
Pb 495274 209115 4195810 7288063 2490956 5485245 6251020 15084254 12003668
Fb 97014 99955 1070424 4621255 705052 1700720 1273878 7599439 42694519
GP 2237600 277516 940676 2184287 1191044 1930426 4414320 2938718 3200810
PP 3237077 6903234 2621056 7929582 437980 516954 4092744 17621829 23680574
HH&SP 5474677 7180750 3561733 10113869 1629023 2447380 8507064 20560548 26881383
2015 Csw 15518623 10366664 7945209 22115615 3480470 31108665 40462151 41238163 19172176
NCsw 140786 80858 1591361 1079113 58869 2628029 1587061 18367543 25047255
Pw 73307 951902 271080 232837 0 3040781 1795945 9348716 46163373
Pb 495426 209000 4202394 7280168 2448638 5429498 6287536 14916988 12576422
Fb 97046 99965 1072798 4660560 691055 1691919 1274196 7617858 43816308
GP 2244675 277559 939740 2183674 1170600 1940740 4435627 2898014 3179639
PP 3231513 6826920 2610425 7922699 435311 517211 4105476 17220457 24236125
HH&SP 5476188 7104480 3550164 10106373 1605911 2457951 8541103 20118472 27415764
2020 Csw 15456829 10649399 8193185 22660135 3591802 31625195 41649087 42709744 19597124
NCsw 141717 81242 1630428 1108924 62793 2685856 1632086 18840167 25592085
Pw 73344 946500 271735 233621 0 3043305 1798315 9323078 46754444
Pb 492986 208844 4225499 7288831 2394629 5375646 6312535 14919615 13128448
Fb 96833 100112 1076199 4714963 624068 1690251 1274134 7644715 44879188
GP 2177970 278563 941347 2188777 1151904 1954488 4495804 2879520 3156562
PP 3055120 6688678 2609079 7997080 431531 517164 4136956 17060385 25014886
HH&SP 5233090 6967241 3550426 10185857 1583436 2471652 8632760 19939905 28171448
2025 Csw 14872282 10788984 8361348 23165713 3676756 32613585 43168636 44591045 19738785
NCsw 140824 81113 1674163 1132742 65296 2752091 1666337 19182850 26126141
Pw 73205 939193 272180 234033 0 3049225 1798960 9326797 47573242
Pb 493707 208330 4252207 7289154 2390027 5305605 6282694 15156731 13687471
Fb 96864 100076 1077963 4775477 607504 1685641 1273474 7783891 46165785
GP 5299748 6863950 3566658 10262072 1577570 2480779 8702434 19802305 29116289
PP 5297963 4882189 4402852 11721879 1673043 2782419 3359900 61927002 71077891
HH&SP 319066 143112 659400 1238082 741151 306517 424868 6634601 9272372

  Enlarge/Reduce  Open in Viewer

Fig. 2 shows the development of the age class distribution of the growing stock when timber assortments are accounted for by the EFDM. The most notable pattern is the increase in volume in the age classes between 40 and 70 years, which results from a changed area distribution over age-classes and an overtime changing stocking level in the separate age-classes.

Fig. 2 - Age class distribution of the growing stock (million m3) after each simulation step. Green color denotes pine, yellow color denotes spruce, and orange color denotes broadleaved species.

  Enlarge/Shrink   Download   Full Width  Open in Viewer

Tab. 6 compares growing stock development when harvests are applied considering timber assortments with the one obtained when the allocation of harvesting activities only accounts for tree species (coniferous and non-coniferous, respectively) and not for timber assortments. The growing stock development differs, mainly as regards coniferous tree species (pine and spruce).

Tab. 6 - The development of growing stock (million m3) by tree species when the harvests are applied with division into timber assortments and without this division.

Standing volumes - With timber assortments Standing volumes - Without timber assortments
Year Pine Spruce Broadleaves Sum Year Pine Spruce Broadleaves Sum
2010 1280 1350 375 3006 2010 1280 1350 375 3006
2015 1429 1341 449 3219 2015 1377 1399 448 3224
2020 1583 1350 530 3464 2020 1480 1458 530 3468
2025 1780 1329 627 3737 2025 1606 1526 624 3756

  Enlarge/Reduce  Open in Viewer

Differences in growing stocks are more pronounced for older age classes, but occur also in the first age classes, for which the effects of final felling and regeneration are visible (Fig. 3). Pines are harvested to a lesser, and spruce to larger, extent when timber assortments are considered. Also for broadleaf species there are, albeit less apparent, differences between the two runs.

Fig. 3 - Differences (million m3) in volume distribution over age classes.

  Enlarge/Shrink   Download   Full Width  Open in Viewer

  Summary and conclusions 

The current study introduces a novel approach for modeling the interaction between wood-based products markets and forest resources, with the aim of providing policy support on a pan-European scale. Specifically, the feedback between a forestry dynamics model (EFDM) and an economic model of the global forest-based sector (GFTM) is elaborated. Besides updating in each period the harvest potential provided by EFDM, contingently upon the demand for primary wood products in GFTM, the suggested framework also takes into account timber assortments when allocating harvest activities.

The outcome of modeling runs indicates that the suggested approach is apt for an integrative modeling of forest resources and wood-based products markets in a pan-European setting. The results also indicate that timber assortments should be considered when modeling this interaction. Hence, accounting for timber assortments in the allocation of harvesting activities over the state-space of EFDM results in different forest states with respect to the case wherein only tree species (coniferous and non-coniferous, respectively) are considered. Further, it also results in different projections of production and trade (and consequently also apparent consumption) of wood-based products. This underscores the importance of a feedback as complete as possible between the forest and the market model.

The EFDM is still under development and a novel method to include also uneven-aged forests within EFDM has recently been developed, and it is currently being tested in a number of countries ([17]). If the case studies prove the validity of the method, the next step in developing the proposed framework will certainly be the expansion of the model coverage, including, besides an increased geographical scope as regards even-aged forests, also short-rotation and uneven-aged forests. In these instances, accounting for timber assortments when linking to a forest-products market models is even more crucial. For example, constants derived from historical removals data used to divide roundwood into assortments are invalidated when applied to timber resulting from drastically different management regimes.

As mentioned above, the allocation of timber potentials and harvests into timber assortments is non-trivial. Ideally, there would be coefficients for each different management activity and forest development state (volume and age class combination) of each forest type. These could be derived from NFI data, but the actual supply of timber assortments from the forest to the markets differs from the theoretical distribution. This information is not recorded by the NFIs, but possibly by the wood-based products industry, or by the contractor conducting harvesting operation. The theoretical coefficients should then be adjusted according to this local, industry-based information. Since these data were not available, in the current analysis we were forced to use the same theoretical coefficients throughout. However, the objective of the study is to construct and demonstrate a well-grounded framework rather than providing quantitative results.

The sustainability criteria used to define the harvest potential of timber assortments were set by means of growing stock and harvest level. The UN General Assembly defines sustainable forest management as a “dynamic and evolving concept, which aims to maintain and enhance the economic, social and environmental values of all types of forests, for the benefit of present and future generations” ([1]). The EFDM has limitations in the sense of not allowing changes between the forest types or land use in the course of a simulation. Only forests available for wood supply were modeled and, therefore, using these strict production-oriented parameters as sustainability criteria is somewhat justified.

In this particular, narrow meaning, sustainability was adhered to, since both the standing volume of forests (Tab. 6) and the annual harvest potential (Tab. 2) increased during the simulation. In addition, the distribution of standing volume covers the entire age distribution. When assessing the supply potential a standard management pattern was used. A larger potential could be found by using more adapted management patterns, but whether this would increase the information content of the potential harvest level is debatable. When meeting the demand for different assortments from GFTM, harvests are allocated over sub-regions and species so as to result in the demanded assortment mix, which could imply that management applied in the simulation steps differs substantially from the program used when assessing the potential.

In the default version of the GFTM, the timber supply potential is split into timber assortments based on historic industrial roundwood removals data from FAOSTAT. Ingesting from the forest resource model the timber supply already allocated into assortments enriches the forest-based sector model. Hence, the market implications of a change in forest management, with an ensuing change in the assortment composition of the timber supply potential, can be modeled with greater detail and realism. Likewise, accounting for timber assortment when allocating harvesting activities in the forest resource model makes it possible to model impacts on forest resources of changing market conditions with greater precision.

Once more, it should be emphasized that in this paper we have deliberately refrained from trying to interpret the results in “real-world” terms, mainly as input data and parameters are still being refined. The purposes of this study is mainly to demonstrate the modeling concept and to assess the linkages between the models. As such, the outcomes of the modeling exercise should be considered with some caution. Hence, they should for example not be seen as detailed forecasts of forest industry developments. Still, already at this stage the modeling approach can be used for scenario analysis, studying directions of change and patterns of causality.

This paper presents a novel concept to model interactions between forest resources and wood-based commodity markets. Of course, there are some aspects that need further elaboration and refinements and we have tried to highlight some of them. Hence, the study has been carried out modeling only one country in the EFDM, to demonstrate the approach and to facilitate the analysis of the results. The outcome of the exercise encourages us to expand the geographical scope of the framework, including more countries in the EFDM. Indeed, the aim is to cover the entire Europe in the elaborated integration. The collection of data needed to reach this aim in terms of coefficients for timber assortments described above constitutes an interesting issue for future research. Another interesting point for future development concerns tree species. Further in the future, the concept should be developed for uneven-aged and short-rotation forest management.

The modeling results of this study indicate that the division into timber assortments has a species-specific impact on harvest allocation. Thus, the explicit consideration of timber assortments affected the propensity of harvesting spruce and pine respectively. At the moment, for Sweden, EFDM is considering three species groups (pine/spruce/broadleaves), while the GFTM currently recognizes only two (conifers/ non-conifers). Having the same detail as to tree species in both models should enhance the dynamics of the model linkage.

A strength of the proposed modeling set-up is that it can be used for analysis with a pan-European (or even global) scope while still being developed. Hence, depending on the current data availability, it is possible to perform in depth analysis for countries where more detailed data sets exist - e.g., as to the impact of a change in forest management regime and corresponding potential timber assortment mix for the provision of wood-based products and forest resource development in a specific country - at the same time as providing assessment on a more general level, e.g., regarding trade implications of a policy change on a pan-European level. In the latter case, the simplified approach as regards forest resource modeling is used for countries where detailed data sets are missing. An important aspect to highlight is that the modeling framework should be developed in close cooperation with national expertise.

A general problem, one which we share with other forest sector modeling enterprises, is that of poor, or at least uncertain, data quality in a number of different respects; from increment rates and growing stocks to input/output coefficients. Hence, an important aspect to highlight is that the modeling framework should be developed in close cooperation with national expertise. This entails, besides provision and assessment of data, also validation of modeling results. Part of this work could possibly be conducted under the auspices of the Forest Information System for Europe (FISE), recently established to support the EU Forest Strategy ([4]).

  List of abbreviations 

The following abbreviations were used throughout the paper:

  1. GFPM: Global Forest Products Model
  2. EFI-GTM: European Forest Institute Global Trade Model
  3. SSPs: Shared Socioeconomic Pathways
  4. Csl: Coniferous sawlogs
  5. Cpw: Coniferous pulpwood
  6. NCsl: Non-coniferous sawlogs
  7. NCpw: Non-coniferous pulpwood
  8. NFI: National Forest Inventory
  9. Csw: Coniferous sawnwood production
  10. NCsw: Non-coniferous sawnwood production
  11. Pw: Plywood production
  12. Pb: Particle board production
  13. Fb: Fibreboard production
  14. GP: Graphical paper production
  15. PP: Packaging paper production
  16. HH&SP: Household & sanitary paper production

  Conflicts of interest 

The authors declare no conflict of interest. The opinions expressed herein are those of the authors and do not necessarily reflect the views of the European Commission.


A/RES/6298 (2008). Non-legally binding instrument on all types of forests A/RES/6298. Food and Agriculture Organization of the United Nations - Forestry, Rome, Italy, pp.
Online | Gscholar
Buongiorno J, Zhu S, Zhang D, Turner J, Tomberlin D (2003). The global forest products model: structure, estimation, and applications. Academic Press, San Diego, CA, USA, pp. 300.
Online | Gscholar
Böttcher H, Verkerk PJ, Mykola G, Havlik P, Grassi G (2012). Projection of the future EU forest CO2 sink as affected by recent bioenergy policies using two advanced forest management models. GCB Bioenergy 4: 773-783.
CrossRef | Gscholar
EC (2013). A new EU forest strategy for forests and the forest-based sector. Communication from the Commission to the European parliament, the Council, the European Economic and Social Committee and the Committee of the Regions. European Commission, Brussels, Belgium, pp. 17.
Fonseca M (2010). Forest product conversion factors for the UNECE region. Geneva Timber and Forest Discussion Paper 49, UNECE/FAO, Geneva, Switzerland, pp. 34.
Jonsson R (2012). Econometric modeling and projections of wood products demand, supply and trade in Europe; Geneva Timber and Forest Discussion Paper 59, UNECE/FAO, Geneva, Switzerland, pp. 192.
Jonsson R, Rinaldi F, San-Miguel-Ayanz J (2015). The global forest trade model - GFTM. JRC Technical Reports, Publications Office of the European Union, Luxembourg, pp. 45.
CrossRef | Gscholar
Kallio AMI, Moiseyev A, Solberg B (2004). The global forest sector model EFI-GTM - the model structure. EFI Internal Report 15, European Forest Institute, Joensuu, Finland, pp. 24.
Online | Gscholar
Kallio AMI, Moiseyev A, Solberg B (2006). Economic impacts of increased forest conservation in Europe: a forest sector model analysis. Environmental Science and Policy 9: 457-465.
CrossRef | Gscholar
Lindner M, Eggers J, Zanchi G, Moiseyev A, Tröltzsch K, Eggers T (2007). Environmentally compatible bio-energy potential from European forests. European Environment Agency, Copenhagen, Denmark, pp. 53.
Lämås T, Eriksson LO (2003). Analysis and planning systems for multiresource, sustainable forestry: the Heureka research programme at SLU. Canadian Journal of Forest Research 33: 500-508.
CrossRef | Gscholar
Moiseyev A, Solberg B, Kallio AMI, Lindne M (2011). An economic analysis of the potential contribution of forest biomass to the EU RES target and its implications for the EU forest industries. Journal of Forest Economics 17: 197-213.
CrossRef | Gscholar
Nabuurs GJ, Päivinen R, Schanz H (2001). Sustainable management regimes for Europe’s forests - a projection with EFISCEN until 2050. Forest Policy and Economics 3: 155-173.
CrossRef | Gscholar
Packalen T, Sallnäs O, Sirkiä S, Korhonen KT, Salminen O, Vidal C, Robert N, Colin A, Belouard T, Schadauer K, Berger A, Rego F, Louro G, Camia A, Räty M, San-Miguel-Ayanz J (2014). The European forestry dynamics model (EFDM). JRC Scientific and Policy Reports, Publications Office of the European Union, Luxembourg, pp. 16.
Rinaldi F, Jonsson R, Sallnäs O, Trubins R (2015). Behavioral modeling in a decision support system. Forests 6: 311-327.
CrossRef | Gscholar
Sallnäs O (1990). A matrix growth model of the Swedish forest. Studia Forestalia Suecica 183: 23.
Online | Gscholar
Sallnäs O, Berger A, Räty M, Trubins R (2015). An area-based matrix model for uneven-aged forests. Forests 6: 1500-1515.
CrossRef | Gscholar
Samuelson PA (1952). Spatial price equilibrium and linear programming. American Economic Review 42: 283-303.
Online | Gscholar
Schelhaas MJ, Didion M, Hengeveld G, Nabuurs GJ, Mason B, Lindner M, Moiseyev A, Edwards D, Jay M, Jensen F, Lucas B, Marzano M, Montagne C, Peace A, Weiss G (2010). Deliverable D2.3.7. Project EFORWOOD “Modeling the impacts of forest management alternatives on recreational values in Europe”, Forest Research, Roslin, UK, pp. 19.
Online | Gscholar
Solberg B, Moiseyev A, Kallio AMI (2003). Economic impacts of accelerating forest growth in Europe. Forest Policy and Economics 5: 157-171.
CrossRef | Gscholar
Solberg B, Moiseyev A, Kallio AMI, Toppinen A (2010). Forest sector market impacts of changed roundwood export tariffs and investment climate in Russia. Forest Policy and Economics 10: 17-23.
CrossRef | Gscholar
UN (2011). European forest sector outlook study II. United Nations, FAO/UNECE Agriculture and Timber Division, Timber Branch, Geneva, Switzerland, pp. 107.
Verkerk PJ, Anttila P, Eggers J, Lindner M, Asikainen A (2011a). The realisable potential supply of woody biomass from forests in the European Union. Forest Ecology Management 261: 2007-2015.
CrossRef | Gscholar
Verkerk PJ, Lindner M, Zanchi G, Zudin S (2011b). Assessing impacts of intensified biomass removal on deadwood in European forests. Ecological Indicators 11: 27-35.
CrossRef | Gscholar

Authors’ Affiliation

Ragnar Jonsson
Francesca Rinaldi
Minna Räty
Ola Sallnäs
European Commission, Joint Research Centre (JRC), Institute for Environment and Sustainability (IES), Forest Resources and Climate Unit, v. E. Fermi 2749, I-21027 Ispra (Italy)

Corresponding author

Ragnar Jonsson


Jonsson R, Rinaldi F, Räty M, Sallnäs O (2016). Integrating forest-based industry and forest resource modeling. iForest 9: 743-750. - doi: 10.3832/ifor1961-009

Academic Editor

Luca Salvati

Paper history

Received: Dec 21, 2015
Accepted: Jul 01, 2016

First online: Aug 12, 2016
Publication Date: Oct 13, 2016
Publication Time: 1.40 months

© SISEF - The Italian Society of Silviculture and Forest Ecology 2016

  Open Access

This article is distributed under the terms of the Creative Commons Attribution-Non Commercial 4.0 International (https://creativecommons.org/licenses/by-nc/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.

Creative Commons Licence

Breakdown by View Type

(Waiting for server response...)

Article Usage

Total Article Views: 7276
(from publication date up to now)

Breakdown by View Type
HTML Page Views: 4569
Abstract Page Views: 204
PDF Downloads: 1922
Citation/Reference Downloads: 64
XML Downloads: 517

Web Metrics
Days since publication: 1158
Overall contacts: 7276
Avg. contacts per week: 43.98

Article citations are based on data periodically collected from the Clarivate Web of Science web site
(last update: Aug 2019)

Total number of cites (since 2016): 4
Average cites per year: 1.00


Publication Metrics

by Dimensions ©

List of the papers citing this article based on CrossRef Cited-by.


iForest Similar Articles

iForest Database Search

Google Scholar Search

PubMed Search


This website uses cookies to ensure you get the best experience on our website