1. Forest simulation models in Norway: main developments and challenges Kjell Andreassen and Tron Eid COST ACTION FP0603: Forest models for research and decision support in sustainable forest management 1st Workshop and Management Committee Meeting. Institute of Silviculture, BOKU. 8-9 of May 2008 Vienna, Austria

2. Main features of Norwegian forests Forest cover (total/share): 8 mil. ha / 28 % of land = productive forest 2 mil. ha / 6 % = Non-productive forest 13 mil. ha / 46 % = Higher altitude than coniferous forest boundary line Growing stock, annual growth and cuts: 680 mil. m 3, 25 mil m 3 y -1, ca 35% of the annual growth is cut Main species: Picae abies (46%), Pinus sylvestris (32%), Betule pendula/pubescens (14%), others (8%) Main non-wood products and services: Berries, hunting (35.ooo moose hunted/yr), Climate restrictions/protection, biodiversity, Recreation, cabins Main risks: Droughts Windthrows Fungi & insect attacks Water floods Effects of climate change Few forest fires (0.004% of area/yr => 23500 yr rotation of fires) Management and silvicultural characteristics: Clear cutting most used for Norway spruce (10 000 ha/yr planted) Seed tree method for Scotch pine “Living forest” standards for sustainable forest management Very little other/alternative forest management methods High value of some non-timber products and services – cf. moose hunting and cabins

3. Forest modelling approaches and trends Empirical models Stand models (“mean tree method”) are widely used (i) The trend has been towards individual tree-level (ii) Models exist for the main coniferous trees, birch and “other broadleaves” Diameter distribution models for the main species Recent research has been concentrated to: New growth models has been developed Modelling regeneration Modelling uneven-aged management Modelling non-timber products and services

4. Forest simulator

5. Forest modeling approaches and trends in Norway SystemBiologyEconomyMulti- functional aspects User flexibility S50-software (Gotaas 1967)Volume directly -NoNone Avvirk1 (Hobbelstad 1979)Volume directly -NoA little Avvirk2 (Hobbelstad 1981)Volume Directly -NoSome Avvirk3 (Hobbelstad 1988)“Average” Tree yesNoGood Gaya-JLP (Hoen & Eid 1990)“Average” Tree yesNoHigh Bestprog (Blingsmo & Veidahl 1991)“Average” Tree yesNoGood Avvirk-2000 (Eid & Hobbelstad 1999  “Average” Tree yesSomeGood Gaya/Sgis (Hoen & Gobakken 1997  “Average” tree yesYesHigh T (Gobakken, Lexerød & Eid 2007  Individual tree

6. Advantage of Individual-tree models vs Stand level based models May describe forest dynamics according to: - selective cutting in un-even-aged forest - different thinning methods, seed trees and shelter wood in even- aged forest - traditional treatments in even-aged forest Timber quality (tree size and wood properties) Forest structure (within-stand diversity) Multi-functional aspects (e.g. lichen, biomass/carbon) Permanent country representative sample plots from NFI Laser-scanning -> diameter distribution -> individual trees

7. Sub-models used in the T-simulator TaskDep. var. Independent variables TreeStandSite Recruitment young even- aged forest (Lexerød & Eid 2005) iN iN A, N, PN sp ALT, LAT, SI Height growth young even- aged forest (Tveite 1977) hASI Recruitment old even-aged and uneven- aged forest (Lexerød 2005) iNiN BA, PBA sp, N, H D, LAY ALT, LAT, SI, VT Diameter growth old even-aged and uneven-aged forest (Bollandsås 2008) idid d, balBA, PBA sp SI, LAT, R Height relationship old even-aged and uneven-aged forest Bollandsås (2008) hdALT, LAT, SI Mortality old even-aged and uneven- aged forest (Eid & Tuhus 2001) Psd, balPBA sp SI

8. Economical sub-models Timber price =f (dbh, h) Cost functions for harvesting/forwarding =f(d bh, volume ha-1, harvesting method)

9. Modelling non-timber products and services Moose hunting (meat and recreation) vs timber production. Important in Norway Recreation in forest areas Water yield, water quality, trade-offs of water and forest

10. Models for predicting risk of hazards Windthrows ( some simple models is developed ) Fire ecolocical effects Fire probability and damage : Ecological important for many species. Effects on vegetation, mosses

11. Simulators and information systems Inventory NFI Forest owners assosiation Stand level simulators Bestprog Gaya/Sgis Avvirk 2000 T - simulator Forest level decision support systems Avvirk 2000 Gaya Process based simulators No process based simulators are still developed in Norway

12. Future challenges How to simulate mixed forests and mixed treatment Process based models - conversion from international models Non-timber products and services Model risk and succession (after hazards) Distance dependent individual tree growth models A more flexible stumpage value calculation system Models describing wood properties Multiple use and environmental aspects Other production functions (berries, forest structure) 3-D GIS interface Carbon budget Integrate and demonstrate the potential of models and simulators for both scientists and other users. Further development/calibration of basic biological models.

13. Some references Eid, T. & Hobbelstad, K. 2000. AVVIRK-2000 - a large scale forestry scenario model for long-term investment, income and harvest analyses. Scand. J. For. Res. 15: 472- 482. Eid, T. 2004. Testing a large-scale forestry scenario model by means of successive inventories on a forest property. Silva Fenn. 3: 305-317. Lexerød, N. & Eid, T. 2006. Assessing suitability for selective cutting using a stand level index. Forest Ecology and Management 237:503-512. Gobakken, T., Lexerød, N. & Eid, T. 2008. T – a forest simulator for bio-economic analyses based on models for individual trees. In press. Scandinavian J. of Forest Research. Andreassen, K. & Tomter, S. 2003. Basal area growth models for individual trees of Norway spruce, Scots pine, birch and other broadleaves in Norway. Forest Ecology and Management 180:11-24.