Forest Inventory Components of Change Growth, Removals

1
Forest Inventory Components of Change (Growth,
Removals Mortality)
Where do they come from?

• Christopher M. Oswalt
• Southern Research Station FIA

2
Outline

• Brief introduction of the topic
• Walk through the components of change
• Examples
• Recent regional trends
• Challenges
• Summary

3
Introduction

• Development informs utilization
• Of interest
• A population as recruitment occurs into, through,
  and eventually exiting some expressed standard

4
Conceptually Simple
Inventory Then
G
M
G
G
G
G
G
G
R
R
R
G
G
G
G
G
G
M
M
R
Inventory Now
5
Operationally Complex
6
Components of Change

• Growth on ingrowth
• Mortality growth
• Cut growth
• Reversion growth
• Diversion growth
• Cull increment growth
• Cull decrement growth

• Survivor growth
• Ingrowth
• Mortality
• Cut
• Reversions
• Diversions
• Cull increment (minus)
• Cull decrement

7
Assumptions

• Time of change is between then and now, but
  exactly when is unknown
• Time of change is estimated to be midway between
  t (time 1) and t1 (time 2)
• Growth model is used to predict midpoint tree
  size
• Usually based on past attributes
• Based on current attributes when there are no
  past attributes

8
Components of GRM

• Survivor growth
• Trees alive and measured time 1 and time 2
• Diameter gt threshold (e.g. 5.0) both times
• No change in tree class (growing stock, cull,
  etc.)
• Contributes to G

Threshold DBH
9
Components of GRM

• Ingrowth
• Tree below threshold time 1 (may or may not have
  been measured)
• Tree grows across threshold before time 2
• Contributes to G

Threshold DBH
10
Components of GRM

• Mortality
• Tree alive time 1
• Tree dead time 2
• Predicted midpoint diameter used to compute
  volume
• Contributes to M

11
Components of GRM

• Cut
• Tree alive time 1
• Tree cut before time 2
• Predicted midpoint diameter used to compute
  volume
• Contributes to R

12
Components of GRM

• Cull increment
• Growing stock time 1
• Cull time 2
• Predicted midpoint diameter used to compute
  volume
• Reduces G

13
Components of GRM

• Cull decrement
• Cull time 1
• Growing stock time 2
• Predicted midpoint diameter used to compute
  volume
• Contributes to G

14
Components of GRM

• Reversion
• Nonforest (or non-timberland) time 1
• Forest (or timberland) time 2
• Predicted midpoint diameter used to compute
  volume
• Contributes to G

15
Components of GRM

• Diversion
• Forest (or timberland) time 1
• Nonforest (or non-timberland) time 2
• Predicted midpoint diameter used to compute
  volume
• Contributes to R

16
GRMs Defined (Scott et. al 2005)

• Growth
• Gross ingrowth Ingrowth Reversion
• Gross growth Gross ingrowth accretion
• Accretion GSGIGRGMGCGD
• Net growth Gross growth - mortality
• Removals
• Cut stems and/or Diversions
• (forest nonforest or timberland-reserved)
• Mortality
• Death from natural causes

17
Growth Removal Trends
Growing Stock on Timberland
Solid line - Growth Broken line - Removals
18
Growth Removal Trends
Growing Stock on Timberland
19
Challenges

• Historical reporting biases and naming
  conventions
• Not all FIA reported removals reflect removed
  trees
• Change and growth not the same thing
• Address through additional tables or
  definitional changes?
• Historic variability in interpreting reserve
  status
• GRM estimates are highly sensitive to changes in
  reserve status
• Address from what angle - data acquisition or
  compilation/processing?
• Not applicable on forest land
• Regional inconsistencies
• Current estimation procedures
• Mapped plot to mapped plot implementation

20
Additional Detail
Total AAR AAR(diversions) More informative
picture of actual removals
21
Challenges

• Historical reporting biases
• Not all FIA reported removals reflect removed
  trees
• Address through additional tables or
  definitional changes?
• Historic variability in interpreting reserve
  status
• GRM estimates are highly sensitive to changes in
  reserve status
• Address from what angle - data acquisition or
  compilation/processing?
• Not applicable on forest land
• Regional inconsistencies
• Current estimation procedures
• Mapped plot to mapped plot implementation

22
Currently Not all there yet.

• South
• Primarily annual to annual
• North
• Primarily annual to annual
• Rocky Mountain
• Remeasurement has not started
• Pacific Northwest
• Testing in California National Forests
• Similar plot footprints (periodic to annual) in
  Alaska (testing)

23
Challenges

• GRM labeling conventions to facilitate proper
  interpretation
• FIDO II
• Tennessee 2007 GRM - Tennessee (47) -- Mortality
  growing-stock on timberland by U.S. Counties
  and Major species group (in cuft/year)

2000
2002
2001
2003
2005
2006
2008
2009
2004
2007
24
Challenges

• Lack of access to previous condition when using
  external data tools
• Internal uses previous condition (correct)
• External uses current condition (wrong)

Example TN(2007) Removals of live trees on
forest land by stand size class
Million ft3 year-1
25
Summary

• GRM is conceptually simple
• GRM analysis is complicated by the fact that the
  same tree can contribute to different components
  of GRM
• GRM analysis will become simpler and more precise
  in the future
• Additional analytical techniques needed
• Refinements in methodology potentially necessary
• Updates to some external tools needed

26
Acknowledgements

• Ray Sheffield
• Gary Brand
• Bill Burkman
• John Coulston