CarboScen
Analysis of carbon outcomes in landscape scenarios
Markku Kanninen & Markku Larjavaara
VITRI, University of Helsinki & CIFOR

How much carbon in a landscape in
the future?
 REDD+ policy maker:
• Policy options – impacts
in terms of emissions
and sequestration
 Land-user planner:
• Improving rural
livelihoods – carbon
outcomes of different
schemes
 Investment advisor:
• Carbon outcomes of
investment options
Past
Future 1 Future 2
Present
Future scenarios of
landscapes

How to quantify ecosystem carbon when
land-uses are known?
• When carbon densities (Dc) of a given land use are
known, just multiply these with the area to get the total
carbon pool (Ctot):
• Ctot = Dc x Area
• But there are two additional points to consider
• First, land-use changes constantly
• Second, carbon densities do not change suddenly when
a land-use change happens

CarboScen tool
Carbon analysis at landscape scale
Carbon data for
different land uses
Scenarios of land
use change
Carbon outcomes in
different land use scenarios
Past Present
Future 1 Future 2
In each scenario
 Total carbon
 By land-use
 By pool
 Above-ground
 Below-ground
 Soil
IPCC Carbon pools,
grouped as
 Above-ground
 Below-ground
 Soil

Published in Ecography (open access)

Software available in CIFOR web site

How CarboScen works?
Assumption: Carbon density approaches the new
equilibrium at a rate, which is a fixed proportion of
the remaining difference per unit time
Carbon
density
(Mg/ha)
Time
Land use change

Simplest case: two land-uses
Cropland
ForestForest
Forest
Cropland
Total C pool in biomass
Cropland
Total C pool in biomass and soil
Total C pool in soil
(Mg/ha)
(ha)
(Mg) (Mg)/area

From where to get carbon density data?
Best option – local data
Additional data collection might be needed
Literature, global data banks

Parameter for the transition rate (rate of
change) of carbon density is more
challenging to obtain
Transition rate (parameter f):
Is the proportion of the remaining
difference between the current carbon
density and the future equilibrium carbon
density that happen annually
Global data needed for most landscapes

Accumulation of
above-ground
biomass with
increasing stand
age
In the Appendix of the
Ecography article
Larjavaara et al. 2017

The transition rate of soil carbon density
is even more challenging
Table: Values of parameter f with standard deviations,
coefficient of variation, and the number of
chronosequences for each conversion type
Conversion type f SD CV n
Cropland to forest 0.056 0.067 120 7
Forest to cropland 0.073 0.055 75 71
Cropland to grassland 0.073 0.061 84 3
Grassland to cropland 0.151 0.149 99 3
Average 0.074 0.061 82 84

Soil carbon pool after conversion from
forest to pasture in Costa Rica
SOC (Mg ha-1) after
conversion
(Veldkamp, 1994)

Soil carbon pool after conversion -
CarboScen

Existing data on soil
carbon density
transition rate is
messy
Few data and high uncertainty
(Poeplau et al. 2011)

Carbon pools after converison – 2
cases - CarboScen

CarboScen –
hands-on training

CarboScen – hands-on
Time Activity
30 minutes • Introduction
• Starting CarboScen
• Form pairs
• Use your own laptop computers
30 minutes • Decide where the landscape for your simulation is
• Decide what are the land use classes (2-4 is the recommended
number for this short exercise)
30 minutes • Search data on biomass carbon densities (e.g. Mg/ha) of your land
use classes
• Decide the depth to which soil carbon is included and search data on
soil carbon densities
• Set transition rates by guessing and comparing to your experience or
by checking the appendices of the Ecography paper on CarboScen
30 minutes • Run at least two realistic or unrealistic simulations and sabe the
results
• Prepare a figure on the most interesting result and get ready to
present that
30 minutes • Presentations by participants

Central Kalimantan (KFCP)
Biomass carbon
density at equilibrium
(Mg ha-1)
Soil carbon density
at equilibrium
(Mg ha-1)
Peat forest (canals blocked, zero logging) 367 5024
Peat forest (some logging) 220 4773
Peat forest (severe logging) 149 4295
Peat shrub land 41 3014
Peat fern or grass 14 3014
Peat oil palm 90 2512
Riparian forest 243 113
Riparian agroforest 102 113
Riparian cropland or settlement 43 68

Central Kalimantan (KFCP)
Biomass carbon
density at equilibrium
(Mg ha-1)
Soil carbon density
at equilibrium
(Mg ha-1)
Restored forest on peat 249 5538
Forest on peat 166 2077
Oil palm plantation on peat 78 1869
Degraded burned peatland 41 1246
Forest on mineral soil 204 113
Non-forest on mineral soil 74 102
Oil palm plantation on mineral soil 78 81
Degraded land on mineral soil 41 81

Simulations based
on workshops

CarboScen – Application 1
• Based on participatory multi-stakeholder
workshops to develop divergent future scenarios
• 4 continents, 4 countries, 8 landscapes and 8
workshops
• In each workshop 4 groups who were given task
to visualize on map the future of the landscapes
with changing drivers of land-use change

Location of the landscapes

Published in Environmental
Research Letters (open access)

Landscape in
Central
Kalimantan

Scenarios

Results on total ecosystem carbon
(Mg/ha) for the four scenarios
0
500
1000
1500
2000
2500
3000
3500
4000
4500
5000
2015 2035 2055 2075 2095 2115
1
2
3
4

Landscape in
West
Kalimantan

Scenarios

Results on total ecosystem carbon
(Mg/ha) for the four scenarios
0
500
1000
1500
2000
2500
2014 2024 2034 2044 2054 2064 2074 2084 2094 2104 2114
1
2
3
4

Carbon simulations
based on interviews

• How much does increasing ecosystem carbon cost?
• Various research methodologies exist
• Why haven’t we asked from the best experts?
• Interviews around the world with the same
methodology allow direct comparisons
• Still not published – minor revision submitted to
Nature Climate Change
CarboScen – Application 2

Interview questions
• Compared to the reference or business-as-usual
scenario
• How would land use change with an annual payment
of USD1 and USD10 paid for every additional metric
ton of ecosystem carbon
• Assumptions:
• Payments are for an infinite period of time
• From a global fund
• Central government decides on mechanisms of
payment
• Good governance

Same landscapes but with Finland

Year
2015 2045 2075
Carbondensity(Mgha
-1
)
100
500
5000
1000
FinlandNorth
FinlandSouth
IndonesiaEast
IndonesiaWest
PeruNorth
PeruSouth
TanzaniaWest
TanzaniaEast
50
MexicoEast
MexicoWest

Study Site Carbon
density
2015
Additional carbon
density - payment
of 10 USD
FinlandNorth 130 15
FinlandSouth 109 9
IndonesiaEast
(Central Kalimantan)
4608 150
IndonesiaWest
(West Kalimantan)
1934 111
MexicoEast 151 7
MexicoEWest 95 3
PeruNorth 160 9
PeruSouth 166 10
TanzaniaEast 80 3
TanzaniaWest 64 10

Additionalrelativetoinitialcarbon
0,0
0,1
0,2
0,3
Fin-
land
North
Fin-
land
South
Indone-
sia
East
Indone-
sia
West
Peru
South
Tan-
zania
East
Tan-
zania
West
Mex-
ico
East
Mex-
ico
West
Peru
North

Biodiversity simulations

• Biodiversity in same landscape and some
additional ones
• Assumed that change in “biodiversity conservation
values” is as fast as the speed of change of
biomass after a land use change
• Biodiversity conservation values obtained based on
interviews of taxonomists and conservation
biologists
• All kinds of taxa included – both fauna and flora
• Research under progress – Mexico and Peru done
CarboScen – Application 3

Example from MexicoWest
20 experts, 51 taxa of fauna and flora

Correlation between biodiversity and carbon
Biomass C Soil C Total C
Biodiversity conservation value
0
10
20
30
40
50
60
70
80
0 5 10
0
20
40
60
80
100
0 5 10
0
20
40
60
80
100
120
140
160
0 5 10

Impact of carbon payment on biodiversity
Biodiversity

Thank you
markku.kanninen@helsinki.fi
markku.larjavaara@gmail.com