Skip to content

JohnKilbride/westfall-2006

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

84 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Westfall & Laustsen (2006): Height-to-Diameter Model for Maine Tree Species

Python License: MIT Forest Science

Poplar height vs. diameter

Predicted height vs. diameter at breast height (DBH) for a hypothetical poplar using the allometric model from Westfall and Laustsen (2006).

Overview

This package provides predict_height_westfall, a function that predicts tree height (ft) at a given top diameter using the Chapman-Richards allometric model from Westfall and Laustsen (2006).

It supports 18 species groups in Maine and accepts diameter at breast height (DBH, inches), compacted crown ratio (percent), tree class, crown class, and an optional top diameter (inches).

Setting top_diam_in=0 (the default) returns total tree height; providing a non-zero top diameter returns the height to that stem diameter, enabling estimation of merchantable bole height.

All six parameters accept either a scalar or an array-like, so mixed-species, mixed-class stands can be predicted in a single call. When any parameter is array-like the inputs are broadcast together and a NumPy array is returned; otherwise a single float is returned.

Model description

Here, we show Equation (2) from Westfall and Laustsen (2006) modified to remove the error term and the random-effects parameters. The model is based on the Chapman-Richards growth equation (Richards 1959).

$$ \begin{aligned} H_{i} = & \left(\beta_0 D_{i} + \beta_1 CC_{1i} + \beta_2 CC_{2i} + \beta_3 CC_{3i}\right) \\ & \cdot \left(1 - \exp\left(-\beta_4 DBH_i\right)\right)^{\beta_5 CR_i + \beta_6 TC_i + \left(\frac{D_{i}}{DBH_i} + 0.01\right)^{\beta_7}} \end{aligned} $$

where:

$H_{i}$ = tree height (feet) of the $i^{th}$ tree at top diameter $D$.

$D_{i}$ = top-diameter (inches) within the $i^{th}$ tree.

$DBH_i$ = diameter at breast height (inches) of the $i^{th}$ tree.

$CC_{1i}$, $CC_{2i}$, $CC_{3i}$ = The crown class indicators of the $i^{th}$ tree. These are a set of one-hot encoded variables.

$$ CC_{ki} = \begin{cases} k = 1, ; = 1 & \text{intermediate, dead; 0 otherwise} \\ k = 2, ; = 1 & \text{dominant, codominant, open grown; 0 otherwise} \\ k = 3, ; = 1 & \text{overtopped; 0 otherwise} \end{cases} $$

$TC_i$ = The tree class of the $i^{th}$ tree. TC is an integer coded value between 1 and 3.

$$ TC_i = \begin{cases} 1 & \text{preferred} \\ 2 & \text{acceptable} \\ 3 & \text{rough/rotten cull, dead} \end{cases} $$

$CR_i$ = compacted crown ratio (%; integers 0 - 100) of the $i^{th}$ tree.

$\beta_0 \text{--} \beta_7$ = the fixed-effects population parameters.

predict_height_westfall

predict_height_westfall(
    species_group: int | array_like,
    dbh_in: float | array_like,
    ccr_pct: float | array_like,
    tree_class: str | array_like,
    crown_class: str | array_like,
    top_diam_in: float | array_like = 0.0,
    fia_spcd: int | array_like | None = None, # OPTIONAL
) -> float | numpy.ndarray

Predicts tree height (ft) at a specified top diameter using the Chapman-Richards allometric model.

Parameter Type Description
species_group int or array Species group number (1–18). Mutually exclusive with fia_spcd; exactly one must be provided.
dbh_in float or array Diameter at breast height (inches, > 0).
ccr_pct float or array Compacted crown ratio (percent, 0–100).
tree_class str or array One of "preferred", "acceptable", "rough", "rotten", or "dead".
crown_class str or array One of "dominant", "codominant", "intermediate", "overtopped", "open grown", or "dead".
top_diam_in float or array Top stem diameter (inches, >= 0) at which to predict height. Defaults to 0.0 for total tree height.
fia_spcd int or array FIA species code(s) (keyword-only). Converted to species group numbers before prediction. Mutually exclusive with species_group; exactly one must be provided.

Returns: Predicted height in feet as a float (scalar inputs) or numpy.ndarray (array inputs).

Species Groups

The species_group parameter accepts integers 1–18. The table below lists the species group number, the group name, the species associated with the group, and the FIA species codes (SPCD) that correspond to each of the species.

Group Species Group Name Species Names FIA SPCD
1 Miscellaneous softwood Larch (introduced), Norway spruce, Jack pine, Red pine, Pitch pine, Pond pine, Scotch pine 6212, 91, 105, 125, 126, 128, 130
2 Tamarack (native) Tamarack (native) 71
3 Eastern white pine Eastern white pine 129
4 White spruce White spruce 94
5 Black spruce Black spruce 95
6 Red spruce Red spruce 97
7 Balsam fir Balsam fir 12
8 Eastern hemlock Eastern hemlock 261
9 Northern white-cedar Northern white-cedar 241
10 Sugar maple Sugar maple 318
11 Ash White ash, Black ash, Green ash 541, 543, 544
12 Poplars Balsam poplar, Eastern cottonwood, Bigtooth aspen, Swamp cottonwood, Quaking aspen 741, 742, 743, 744, 746
13 Miscellaneous hardwood Shagbark hickory, Black cherry, Scarlet oak, Northern red oak, Black oak, American basswood 407, 762, 806, 833, 837, 951
14 Yellow birch Yellow birch 371
15 Paper birch Paper birch 375
16 Other hardwood Maple, Striped maple, Silver maple, Mountain maple, Norway maple, Ohio buckeye, Serviceberry, Sweet birch, Gray birch, American hornbeam, Butternut, Osage-orange, Apple, Eastern hophornbeam, Pin cherry, Chokecherry, White oak, Swamp white oak, Willow, Black willow, White willow, American mountain-ash, Elm, American elm 310, 315, 317, 319, 320, 331, 356, 372, 379, 391, 601, 641, 660, 701, 761, 763, 802, 804, 920, 922, 927, 935, 970, 972
17 Red maple Red maple 316
18 American beech American beech 531

Demonstration calculations

Westfall and Laustsen (2006) provided the following example of how to apply the equations using a poplar tree (species group = 12) with the following attributes:

  • dbh = 15.5 in.
  • Compacted crown ratio = 40 percent
  • Tree class = acceptable (TC = 2)
  • Crown class = codominant (CC₁ = 0; CC₂ = 1; CC₃ = 0)

We demonstrate that this implementation reproduces the same results.

1) Prediction of total height

When computing total height, you set the top-height diameter to 0.

$$\begin{aligned} H_{i0} &= \left(-4.2401(0) + 84.2529(0) + 91.5048(1) + 78.7788(0)\right) \\ &\quad \cdot \left(1 - \exp(-0.1023 \cdot 15.5)\right) \\ &\quad \cdot \exp\left(0.0054(40) + 0.0638(2) + \left(\tfrac{0}{15.5} + 0.01\right)^{0.1422}\right) \\ H_{i0} &= (91.5048) \cdot (0.7943) \cdot \exp\left(0.3436 + (0.01)^{0.1422}\right) \\ H_{i0} &= 75.0 \text{ ft} \end{aligned}$$

from westfall_2006 import predict_height_westfall

total_height = predict_height_westfall(
    species_group = 12,
    dbh_in = 15.5,
    ccr_pct = 40,
    tree_class = "acceptable",
    crown_class = "codominant",
    top_diam_in = 0.0,
)
# total_height => 75.0 ft

2) Prediction of bole height (4-in. top diameter)

Compute the top height of the 4-in. diameter bole.

$$\begin{aligned} H_{i4} &= \left(-4.2401(4) + 84.2529(0) + 91.5048(1) + 78.7788(0)\right) \\ &\quad \cdot \left(1 - \exp(-0.1023 \cdot 15.5)\right) \\ &\quad \cdot \exp\left(0.0054(40) + 0.0638(2) + \left(\tfrac{4}{15.5} + 0.01\right)^{0.1422}\right) \\ H_{i4} &= (74.5444) \cdot (0.7943) \cdot \exp\left(0.3436 + (0.2681)^{0.1422}\right) \\ H_{i4} &= 56.9 \text{ ft} \end{aligned}$$

bole_height = predict_height_westfall(
    species_group = 12,
    dbh_in = 15.5,
    ccr_pct = 40,
    tree_class = "acceptable",
    crown_class = "codominant",
    top_diam_in = 4.0,
)
# bole_height => 56.9 ft

3) Prediction of sawlog height (9-in. top diameter)

Compute the top height of the 9-in. diameter bole.

$$\begin{aligned} H_{i9} &= \left(-4.2401(9) + 84.2529(0) + 91.5048(1) + 78.7788(0)\right) \\ &\quad \cdot \left(1 - \exp(-0.1023 \cdot 15.5)\right) \\ &\quad \cdot \exp\left(0.0054(40) + 0.0638(2) + \left(\tfrac{9}{15.5} + 0.01\right)^{0.1422}\right) \\ H_{i9} &= (53.3439) \cdot (0.7943) \cdot \exp\left(0.3436 + (0.5906)^{0.1422}\right) \\ H_{i9} &= 39.8 \text{ ft} \end{aligned}$$

sawlog_height = predict_height_westfall(
    species_group = 12,
    dbh_in = 15.5,
    ccr_pct = 40,
    tree_class = "acceptable",
    crown_class = "codominant",
    top_diam_in = 9.0,
)
# sawlog_height => 39.8 ft

Installation

Install the latest release from PyPI:

pip install westfall-2006

Or clone the repository and install as an editable package:

git clone https://github.com/johnkilbride/westfall_height_to_diameter_2006.git
cd westfall_height_to_diameter_2006
pip install -e .

Citation

If you use this library, please cite Westfall and Laustsen (2006).

@article{westfall2006merchantable,
  title={A merchantable and total height model for tree species in Maine},
  author={Westfall, James A and Laustsen, Kenneth M},
  journal={Northern Journal of Applied Forestry},
  volume={23},
  number={4},
  pages={241--249},
  year={2006},
  publisher={Oxford University Press}
}

About

An implementation of the allometric height-to-diameter models from Westfall & Laustsen (2006) for 18 species groups in Maine.

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors