An Introduction toOSU StreamWood
Mark A. Meleason2, Daniel J. Sobota1, Stanley V. Gregory3
1Washington State University, Vancouver Campus2USDA Forest Service Pacific Northwest Research Station
3Department of Fisheries and Wildlife, Oregon State University
Presentation Outline
I. Model Description
II. Types of Applications
III. Simulation Example
I. Model Description
Model Overview
Model Components
Model Performance
OSU StreamWood predicts… STANDING STOCK of wood
(Breakage, movement, and decay)
MEANS and VARIANCE
(Individual–based Stochastic)
GENERAL trends
Scales: Time – ANNUAL
Space – MULTIPLE REACH
STREAMWOOD
Forest Stream
Tree Recruitment
Tree Growth
Tree Mortality
Log Recruitment
Log Breakage
Log Movement
DecompositionForest Harvest
STREAMWOOD
Tree Recruitment
Tree Growth
Tree Mortality
Log Recruitment
Log Breakage
Log Movement
DecompositionForest Harvest
Forest Stream
Forest Inputs
Forest Gap–Phase Model (w/I SW)JABOWA (Botkin et al., 1972)Individual-based, Monte Carlo
ORGANON and FVS (G&Y models)
User defined
no cut
partial cut
Riparian ZoneHarvestRegime
stream
forestupland
STREAMWOOD
Tree Recruitment
Tree Growth
Tree Mortality
Log Recruitment
Log Breakage
Log Movement
DecompositionForest Harvest
Forest Stream
STREAMWOOD
Tree Recruitment
Tree Growth
Tree Mortality
Log Recruitment
Log Breakage
Log Movement
DecompositionForest Harvest
Forest Stream
directionalfall
randomfall
Tree Fall Regime
stream
forestrandomfall ordirectionalfall
STREAMWOOD
Tree Recruitment
Tree Growth
Tree Mortality
Log Recruitment
Log Breakage
Log Movement
DecompositionForest Harvest
Forest Stream
Tree Entry Breakage
BankfullWidth
A1
Log lengths
C3
A2 B2 B1
In-channel Breakage
Does the log break?residence timetop diameter
If so where?Variations on broken stick modelBreak location related to diameter
Predicted vs. Observed
0
10
20
30
40
5 10 15 >=20
Length class (1-m interval)
% o
f al
l pie
ces
Observed
Simulated
STREAMWOOD
Tree Recruitment
Tree Growth
Tree Mortality
Log Recruitment
Log Breakage
Log Movement
DecompositionForest Harvest
Forest Stream
Chance of Log Movement
Does the log move?
Function of:FLOW (peak annual flow)Number of Key PiecesLength outside of channelLength to bankfull width
Chance of Movement: No Key Pieces, 100% Within
Channel
204060801001 0.8 0.
6 0.4 0.
20
20
40
60
80
100
chance move(%)
flow (rec yrs) lbfw ratio
Distance of Log Movement
If it does move, then how far?
Single negative exponential model
k = average travel distance(units of bank full width)
Assumed independent of piece size and channel characteristics
Distance Moved, Mack Creek
y = 98.258e-0.0094x
R2 = 0.9906
n = 643
0
25
50
75
100
0 150 300 450 600 750Distance moved (m)
% m
oved
bey
ond
STREAMWOOD
Tree Growth
Tree Mortality
Log Recruitment
Log Breakage
Log Movement
DecompositionForest Harvest
Forest Stream
Tree Recruitment
Decomposition
Single negative exponential
Represents microbial decay and physical abrasion
Species-specific aquatic and terrestrial rates
The Value of Models
“Models of course, are never true, but fortunately it is only necessary that they be useful.”
“For it is usually needful only that they not be grossly wrong.”
Box, G. E. P. 1979. Some problems of statistics and everyday life. J. Am. Stat. Assoc. 74: 1-4
Model Performance Evaluation“Truth is the intersection of independent lies”
(Levins1970)
Absolute Tests difficult for most models
Using realistic input parameters: Reasonable agreement with available dataAnd derived characteristics (e.g., log
length frequency distribution)
Sensitivity Analysis: ID critical variables
II. Sample Applications
Vary, riparian width, no-cut width, and upland rotation length
Characterizing variability of wood volume for a given forest type
Forest Basal Area: Standard Run
0
25
50
75
100
0 120 240 360 480 600 720
Time (years)
Ba
sa
l A
rea
(m
2/h
a)
PSME TSHE THPL Total
Forest Plantation Basal Areas
0
25
50
75
100
0 120 240 360 480 600 720
Time (years)
Ba
sa
l a
rea
(m
2/h
a)
Standard run R60 R90 R120
Volume From Plantation Forests
0
50
100
150
200
0 120 240 360 480 600 720
Time (years)
Vo
lum
e (
m3/1
00
m)
Standard run B0R60 B0R90 B0R120
Plantation Forests: 6-m Buffer
0
50
100
150
200
0 120 240 360 480 600 720
Time (years)
Vol
ume
(m3/1
00 m
)
B6R0 B6R60 B6R90B6R120 Standard run
Plantation Forests: 10-m Buffer
0
50
100
150
200
0 120 240 360 480 600 720
Time (years)
Vo
lum
e (
m3/1
00
m)
B10R0 B10R60 B10R90B10R120 Standard run
Plantation Forests: 15-m Buffer
0
50
100
150
200
0 120 240 360 480 600 720
Time (years)
Vo
lum
e (
m3/1
00
m)
Control B15R120 B15R90B15R60 B15R0
Total Volume by Buffer Width
0
50
100
150
200
0 120 240 360 480 600 720
Time (years)
Vol
ume
(m3/1
00 m
)
75-m 40-m 30-m 25-m20-m 15-m 10-m 6-m
Study Conclusions
6-m buffer: 32% of site potential
30-m buffer: 90% of site potential
Plantation forests: maximum 1st cut
0
20
40
60
0 450 900 1350 1800Time (year)
Volu
me
(m3 1
00 m
-1)
1800-yr
Simulated Wood Volume Waihaha Basin, New Zealand
Volume Frequency Distribution
Year 1800, Waihaha, NZ
0
5
10
15
20
25
0 10 20 30 40 50
Wood Volume class ( m3 / 100 m)
Rel
ativ
e F
req
uen
cy 1800-yr
0
50
100
0 30 60Total Volume (m3 100 m-1)
Freq
uenc
y
1800-yr
200-yr
400-yr
600-yr
Cumulative Frequency Volume Distribution Waihaha, NZ
III. Simulation Example
4-reach system using the internal forest model (no harvest activity)
Bank full width = 10 m, length =200 m
Run for 200 years, 100 iterations
Final Thoughts Designed to be flexible
Currently v2 is under construction Includes “StreamLine” – a 1-reach system Imports ORGANON and/or FVS dead tree files
Latest release version on HJA LTER website http://www.fsl.orst.edu/lter/data/tools/models/
Developer: Mark Meleason ([email protected])
Questions?Questions?Questions?Questions?