Aggregated modelling of disturbed and undisturbed ...
1
Aggregated modelling of disturbed and undisturbed rainforest dynamics ! Development of an easily applicable rainforest simulation model (FORREG) to calculate bole volume dynamics with FORREG distinguishes three tree species groups How to acquire the parameters? Implementation of a as an optimization process: fit the model parameters to “characteristic values” of the forest genetic algorithm * Britta Tietjen , Andreas Huth * [email protected] Approach of the FORREG model Aims Logging operation Estimation of the current harvestable bole volume with an estimate function Results - Comparison of the Models Discussion Species group Light demand Growth speed Potential height pioneer tree species high fast 15-25 m intermediate tree species low medium 25-36 m climax tree species low slow >36 m P M P mat P mat P P M P c K V V V V K dt dV × + - × × = , , ) 2 ( 1 C mat C CP IP CP I mat I CP IP IP P V V w w w V V w w w c , , 1 × + - × + - = ) 1 ( ) ( , , C mat C CI II CI I mat I CI II II I I input I V V w w w V V w w w V g a dt dV × + - × + - × × + = ) 1 ( ) ( , , C mat C CC IC CC I mat I CC IC IC C C input C V V w w w V V w w w V g a dt dV × + - × + - × × + = ! Assessment of the impact of different logging strategies on yield and on the ability of the forest to re-grow FORREG reproduces the results of a more complex model (FORMIX3-Q) with coupled differential equations Undisturbed forest growth of three sites References UFZ - UMWELT FORSCHUNGSZENTRUM LEIPZIG-HALLE GMBH IN DER HELMHOLZ-GEMEINSCHAFT with Characteristic values of a forest Forest growth simulated with FORMIX3-Q without inter-species competition Linear estimate of the harvestable volume dependent on the cutting threshold diameter (data calculated by FORMIX3-Q) 0 50 100 150 200 250 300 350 400 0 200 400 600 Time [years] Bole Volume [m³/ha] 0 50 100 150 200 250 300 350 400 450 0 200 400 600 Time [years] Bole Volume [m³/ha] Pioneer - FORMIX Intermediate - FORMIX Climax - FORMIX Pioneer - FORREG Intermediate - FORREG Climax - FORREG Tropical rainforest in the Deramakot Forest Reserve (DFR) in Malaysia Tropical rainforest in the Lambir National Park (LNP) in Malaysia Subtropical moist forest in Paraguay (MFP) Different logging strategies in the DFR Cutting threshold diameter 60 cm, Logging cycle 40 years Cutting threshold diameter 60 cm, Logging cycle 80 years 0 500 1000 1500 2000 2500 0 20 40 60 80 100 Logging Cycle [years] Total Yield [m³/ha] 0 1000 2000 3000 4000 5000 Total Damage [m³/ha] Total yield and damage dependent on the applied logging cycle Glauner, R., T. Ditzer, and A. Huth. 2003. Growth and yield of tropical moist forest for forest planning: an inquiry through modeling. Can. J. Forest Res., 33: 521-535. Huth, A. and T. Ditzer. 2001. Long-term impacts of logging in a tropical rain forest a simulation study. For. Ecol. Manage., 142: 33-51. Pearce, D., F.E. Putz, and J.K. Vanclay. 2003. Sustainable forestry in the tropics: panacea or folly? Forest Ecol. Manag., 172: 229-247. Putz, F.E. and V. Viana. 1996. Biological challenges for certification of tropical timber. Biotropica, 28: 323-330. ! With the help of a genetic algorithm a rainforest model was developed, which is relatively ! The , since only aggregated processes are modelled ! The resulting bole volumes of the complex FORMIX3-Q model and the new FORREG model agree altogether well, this applies to undisturbed as well as disturbed sites ! Differences in results for the Lambir National Park can be explained by the fact that no size classes are modelled in the FORREG model. However, these differences do not significantly change results for different logging strategies (not shown) ! The logging cycle has an important influence on the obtained yield and the caused damage, an optimal logging cycle can be found in terms of yield. easy to parameterise simulation time is short # few, easily acquirable parameters # a short simulation time 0 20 40 60 80 100 120 140 160 0 200 400 600 Time [years] Bole Volume [m³/ha]
Transcript of Aggregated modelling of disturbed and undisturbed ...
Aggregated modelling of disturbed and undisturbed rainforest dynamics
! Development of an easily applicable rainforest simulation model (FORREG) to calculate bole volume dynamics with
FORREG distinguishes three tree species groups How to acquire the parameters?Implementation of a as an optimization process: fit the model parameters to “characteristic values” of the forest
genetic algorithm
*Britta Tietjen , Andreas Huth*[email protected]
Approach of the FORREG model
Aims
Logging operationEstimation of the current harvestable bole volume with an estimate function
Results - Comparison of the Models Discussion
Species group Light demand
Growth speed
Potential height
pioneer tree species high fast 15-25 m intermediate tree species low medium 25-36 m climax tree species low slow >36 m
P
M
Pmat
Pmat
PP
M
P cK
V
V
VV
Kdt
dV×+-××= ,
,
)2(1
Cmat
C
CPIP
CP
Imat
I
CPIP
IPP
V
V
ww
w
V
V
ww
wc
,,
1 ×+
-×+
-=
)1()(,, Cmat
C
CIII
CI
Imat
I
CIII
IIIIinput
I
V
V
ww
w
V
V
ww
wVga
dt
dV×
+-×
+-××+=
)1()(,, Cmat
C
CCIC
CC
Imat
I
CCIC
ICCCinput
C
V
V
ww
w
V
V
ww
wVga
dt
dV×
+-×
+-××+=
! Assessment of the impact of different logging strategies on yield and on the ability of the forest to re-grow
FORREG reproduces the results of a more complex model (FORMIX3-Q) with coupled differential equations
Undisturbed forest growth of three sites
References
UFZ - UMWELTFORSCHUNGSZENTRUM LEIPZIG-HALLE GMBH IN DER HELMHOLZ-GEMEINSCHAFT
with
Characteristic values of a forest
Forest growth simulated with FORMIX3-Q without inter-species competition
Linear estimate of the harvestable volume dependent on the cutting threshold diameter (data calculated by FORMIX3-Q)
0
50
100
150
200
250
300
350
400
0 200 400 600
Time [years]
Bo
leV
olu
me
[m³/
ha
]
0
50
100
150
200
250
300
350
400
450
0 200 400 600
Time [years]
Bo
leV
olu
me
[m³/
ha
]
Pioneer - FORMIX Intermediate - FORMIX Climax - FORMIX Pioneer - FORREG Intermediate - FORREG Climax - FORREG
Tropical rainforest in the Deramakot Forest Reserve (DFR) in Malaysia
Tropical rainforest in the Lambir National Park (LNP) in Malaysia
Subtropical moist forest in Paraguay (MFP)
Different logging strategies in the DFR
Cutting threshold diameter 60 cm,Logging cycle 40 years
Cutting threshold diameter 60 cm,Logging cycle 80 years
0
500
1000
1500
2000
2500
0 20 40 60 80 100
Logging Cycle [years]
To
talY
ield
[m³/
ha
]
0
1000
2000
3000
4000
5000
To
talD
am
ag
e[m
³/h
a]
Total yield and damage dependent on the applied logging cycle
Glauner, R., T. Ditzer, and A. Huth. 2003. Growth and yield of tropical moist forest for forest planning: an inquiry through modeling. Can. J. Forest Res., 33: 521-535.Huth, A. and T. Ditzer. 2001. Long-term impacts of logging in a tropical rain forest a simulation study. For. Ecol. Manage., 142: 33-51.Pearce, D., F.E. Putz, and J.K. Vanclay. 2003. Sustainable forestry in the tropics: panacea or folly? Forest Ecol. Manag., 172: 229-247.Putz, F.E. and V. Viana. 1996. Biological challenges for certification of tropical timber. Biotropica, 28: 323-330.
! With the help of a genetic algorithm a rainforest model was developed, which is relatively
! The , since only aggregated processes are modelled
! The resulting bole volumes of the complex FORMIX3-Q model and the new FORREG model agree altogether well, this applies to undisturbed as well as disturbed sites
! Differences in results for the Lambir National Park can be explained by the fact that no size classes are modelled in the FORREG model. However, these differences do not significantly change results for different logging strategies (not shown)
! The logging cycle has an important influence on the obtained yield and the caused damage, an optimal logging cycle can be found in terms of yield.
easy to parameterise
simulation time is short
# few, easily acquirable parameters# a short simulation time
0
20
40
60
80
100
120
140
160
0 200 400 600
Time [years]
Bo
leV
olu
me
[m³/
ha
]
