An Introduction to OSU StreamWood Mark A. Meleason 2, Daniel J. Sobota 1, Stanley V. Gregory 3 1 Washington State University, Vancouver Campus 2 USDA Forest.

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?