Peder: /* Construction of tree populations from aggregated young stand data */

2010-09-27T06:46:01Z

Construction of tree populations from aggregated young stand data

← Older revision		Revision as of 07:46, 27 September 2010
Line 42:		Line 42:

	...		...

			==References==

Peder at 06:44, 27 September 2010

2010-09-27T06:44:54Z

← Older revision		Revision as of 07:44, 27 September 2010
Line 1:		Line 1:
			Under construction!

	Author: Björn Elfving.		Author: Björn Elfving.
	==Introduction==		==Introduction==

Peder: /* Introduction */

2010-09-27T06:44:12Z

Introduction

← Older revision		Revision as of 07:44, 27 September 2010
Line 1:		Line 1:
	Author: Björn Elfving.		Author: Björn Elfving.
	==Introduction==		==Introduction==
	~~Heureka replaces the former planning systems Hugin and Forest Planning Package, and the growth models from those systems are available in Heureka. However, the~~ default options for growth modelling in Heureka are ~~mainly~~ based ~~on new~~ functions which will be presented here. All growth functions are common for the three applications and predictions can be based on single plot data or data from several plots representing a stand.		The default options for growth modelling in Heureka are based functions which will be presented here. All growth functions are common for the three applications StandWise, PlanWise and RegWise, and predictions can be based on single plot data or data from several plots representing a stand.

	Two stages are distinguished at the growth modelling: the stand establishment period (h<7 m) and the development of established stands (h>7 m). Predictions of stand establishment and development of young stands are based on data from the Hugin young stand survey. For un-stocked areas (in initial inventory data or on areas that are clear-cut during the growth predictions) the expected stocking 12 years after clear-cut in natural regenerations or 12 years after sowing or planting is calculated as a function of site conditions and prescribed regeneration measures. With stocking, regeneration type and site conditions as indicators, data on height- and species distribution in the new stand is either imported from a data-base with measured plots or estimated with functions based on data from this data-base. Probabilities for all trees to become a future crop tree are calculated, and from this the mean height of different species in this population. Those probabilities are also used for stem selection at pre-commercial thinning.		Two stages are distinguished at the growth modelling: the stand establishment period (h<7 m) and the development of established stands (h>7 m). Predictions of stand establishment and development of young stands are based on data from the Hugin young stand survey. For un-stocked areas (in initial inventory data or on areas that are clear-cut during the growth predictions) the expected stocking 12 years after clear-cut in natural regenerations or 12 years after sowing or planting is calculated as a function of site conditions and prescribed regeneration measures. With stocking, regeneration type and site conditions as indicators, data on height- and species distribution in the new stand is either imported from a data-base with measured plots or estimated with functions based on data from this data-base. Probabilities for all trees to become a future crop tree are calculated, and from this the mean height of different species in this population. Those probabilities are also used for stem selection at pre-commercial thinning.

Peder: Created page with 'Author: Björn Elfving. ==Introduction== Heureka replaces the former planning systems Hugin and Forest Planning Package, and the growth models from those systems are available in�?�'

2010-09-27T06:42:48Z

Created page with 'Author: Björn Elfving. ==Introduction== Heureka replaces the former planning systems Hugin and Forest Planning Package, and the growth models from those systems are available in�?�'

New page

Author: Björn Elfving.
==Introduction==
Heureka replaces the former planning systems Hugin and Forest Planning Package, and the growth models from those systems are available in Heureka. However, the default options for growth modelling in Heureka are mainly based on new functions which will be presented here. All growth functions are common for the three applications and predictions can be based on single plot data or data from several plots representing a stand.

Two stages are distinguished at the growth modelling: the stand establishment period (h<7 m) and the development of established stands (h>7 m). Predictions of stand establishment and development of young stands are based on data from the Hugin young stand survey. For un-stocked areas (in initial inventory data or on areas that are clear-cut during the growth predictions) the expected stocking 12 years after clear-cut in natural regenerations or 12 years after sowing or planting is calculated as a function of site conditions and prescribed regeneration measures. With stocking, regeneration type and site conditions as indicators, data on height- and species distribution in the new stand is either imported from a data-base with measured plots or estimated with functions based on data from this data-base. Probabilities for all trees to become a future crop tree are calculated, and from this the mean height of different species in this population. Those probabilities are also used for stem selection at pre-commercial thinning.

Inventory data from young stands are often given in aggregated form as number of trees per hectare and mean height for different species. Then a list of single trees is created for each plot by use of distribution functions

Mean age is solved from species, mean height and site index with functions for mean height development. Age at breast height is also calculated for all trees as well as probabilities for survival and severity of damages by different agencies. Height growth is then estimated with the height development curves and differentiation functions, including effects of damages, for all trees in 5-year steps until mean height exceed 7 m. At this stage the stand is considered as established. All trees are then given a diameter as a function of species, height, stand density and other variables and the continued development is predicted with functions for established stands.

Growth of established stands is predicted with separate functions for basal area growth and height growth. Initial data should contain a tree list with species and diameter for all trees with dbh�?�4 cm. Minimum input variables are also stand age , data on performed thinning, latitude, altitude, field vegetation type, soil moisture and site index according to site factors. Stand age is defined as basal-area weighted total mean age of the main stand (excluding standards). The stand must also be classified as even- or uneven-aged, where the latter is defined by having less than 80 % of the main stand volume within a 20-year age-span.

At start of prediction, if not given in initial data, height and age at breast height are calculated for all trees and a new stand age is calculated on basis of this. Time to reach breast height is calculated with a function that assimilates figures given in a table from the NFI field work instruction that has been practiced since 1983. Volumes are calculated with single-tree functions based on data from the Great Yield Investigation performed in the period 1940-1963. Also biomasses in roots, stem, branches and needles are calculated.

Stand development is predicted with 5-year steps. Basal area growth is calculated both tree-wise with functions for single trees and for the whole stand with a stand-level function. The latter function has the highest precision and is used to calibrate the growth level, while the individual-tree functions distribute growth between trees in a realistic manner.

Height growth is predicted with top-height development functions (site curves). For each tree height after 5 years is estimated as a function of initial height and age. Recruitment of new trees with dbh�?�4 cm in established stands is calculated with in-growth functions. Mortality is estimated as a proportion of the basal area before growth with area-based functions and distributed on single trees with models for mortality probability of single trees. Thinning is distributed on species and trees with the same algorithms that were used in the Hugin system. The thinning response is calculated with specific thinning response functions. Effects of whole-tree utilisation at harvest are also considered. The effect of nitrogen fertilisation is initially given in the form of total response in m³sk/ha as function of dose and stand parameters. Stand growth is predicted with ordinary functions until the predicted volume including fertilisation effect is reached and that stand conditions are then used as stand status when fertilisation effect has ceased, that is at a lower age (the age scale is shifted). The effect of genetic improvement is included by an increased site index. The amount of rot on spruce is estimated and taken into account at the valuation of the yield. Effects of global warming can be predicted by selection among different SWECLIM-scenarios. Correction factors for volume growth are available from tables produced with the mechanistic model BIOMASS. Adjustments are made to site and stand conditions.

Shelter trees, seed trees and retention trees can be selected with the thinning algorithms or with specific rules, and the mortality of those trees is predicted with a specific function. For old trees also an age-related mortality factor acts to avoid that trees become unrealistically old. The break-down of dead trees is followed. Competing effects of retained trees on the new stand are also calculated. The output features are very flexible and the development can be followed on the single-tree level. In this document the various sub-models are superficially described, more detailed (technical) descriptions are found in the Heureka model database.

==Regeneration==

==The Hugin Young Stand Survey: database and functions==

==Construction of tree populations from aggregated young stand data==
In forest inventories, the description of young stands is often given by stand-level data like mean height, total number of stems per hectare and species proportions of the number of stems. When such data is imported to Heureka a reconstruction on the plot level of single trees (species and height) is needed. A set of functions have been developed for this purpose by Kenneth Nyström, based on data from the Hugin Young Stand Inventories. For each plot mean height (H) and the coefficient of variation of tree heights (CVH) were computed, over all trees and per species. Furthermore, the two-parameter Weibull distribution was adapted to describe the height distribution on the plots and the parameters were fit to the data using the maximum-likelihood method. Then the estimated distribution parameters were expressed as functions of H and CVH and estimated simultaneously with seemingly unrelated regression techniques (SUR). The functions are documented in the Heureka model database and are applied as follows.

In the first step the stand type is identified by stem number proportions of different species. Seven stand types are distinguished:
# Pine stands: if p(pine) ≥ 0.7
# Spruce stands: if p(spruce) ≥ 0.7
# Contorta stands: if p(contorta) ≥ 0.7
# Mixed conifer stands: if p(conifer) ≥ 0.7
# Species mixed stands: if p(broadleaves) ≥ 0.4 and p(broadleaves) < 0.7
# Broadleaf stands: if p(broadleaves) ≥ 0.7 and p(leaf-tree ) < 0.5
# Leaf-tree stands: if p(broadleaves) ≥ 0.7 and p(leaf-tree) ≥ 0.5

If mean heights are given per species, coefficients of variation for height and Weibull parameters for the height distributions are calculated for each species. If only the total mean height of the stand is given, the corresponding distribution parameters are calculated as functions of stand type, stand origin (cultivation or natural regeneration), mean height, and species composition.

With the estimated distributions the heights of the specified number of trees are created. The frequency form (f(h)) and the integral form (F(h)) of the Weibull function are as follows:

...

Simulation models - Revision history

Peder: /* Construction of tree populations from aggregated young stand data */

Peder at 06:44, 27 September 2010

Peder: /* Introduction */

Peder: Created page with 'Author: Björn Elfving. ==Introduction== Heureka replaces the former planning systems Hugin and Forest Planning Package, and the growth models from those systems are available in�?�'