Estimation of Carbon Stocks

Alex Grais and Felipe Casarim

 Calculating carbon pools
• After field work has been completed, need to
calculate carbon per hectare in each pool
• For plots:
– Perform calculations needed for each pool
– Scale plot measurements to per hectare
– Average across plots per stratum for each carbon
pool
– Sum across pools
Calculating carbon stocks in key
carbon pools

Above Ground
Live Biomass Dead wood
Above Ground
Live Trees

AG non-tree woody Standing

deadwood
AG non-tree non-woody Litter
Lying deadwood

Belowground Live
Biomass
 Carbon versus Biomass
• Carbon estimated to be a constant proportion
of biomass (e.g. IPCC defaults)
– Live biomass, standing+lying dead wood:
Biomass * 0.47 = Carbon
– Litter: Biomass * 0.4 = Carbon
• OR: project can measure proportion in
laboratory using selection of subsamples
taken
 CALCULATING PLOT AREA
• Trees were measured in different NEST sizes
• Area of a circle = A  r 2
– 4 m radius = 50.3 m2
– 14 m radius = 615.8 m2
– 20 m radius = 1256.6 m2
 CORRECTING PLOT AREA FOR SLOPE
• Example:
◦ Plot 20 m radius, slope 25%
◦ L = 20 * cos (25) = 18.1 true plot radius

Plot Area on 0% slope = π x 20 x 20 = 1256.6 m2

Adjusted Plot Area = π x 20 x 18.1 = 1137.3 m2

 Scaling factor: how many plots per ha?
• Measure trees in different nest sizes
• Need to standardize all measurements to a per-hectare
basis
• 1 ha = 10,000 m2
• How many of each size plot are in 1 ha?
– 4 m radius = 50.3 m2 = 198.8 (10,000 / 50.3)
– 14 m radius = 615.8 m2 = 16.2 (10,000 / 615.8)
– 20 m radius = 1256.6 m2 = 8.0 (10,000 / 1256.6)
• Each tree measured in a nest should be multiplied by
the appropriate scaling factor for its nest size
• Need scaling factors for clip plots and litter plots also
 Live tree carbon
• Use regression or allometric equation to estimate
biomass from DBH
1,400
Biomass = 0.0288*DBH
 Relationship between
2.6948
tree
1,200 R2 = 0.98diameter and mass
1,000
(‘biomass’) of tree
 Many equations published
Biomass (kg C per tree)

800
for forests worldwide
 Local regression equations
600

400
may exist in literature
200

0
0 20 40 60
DBH (cm)
 Live tree carbon
• Calculate biomass of each tree (kg)
• Sum biomass of trees in each nest size
• Multiply biomass by scaling factor for nest size
• Sum biomass/ha of each nest for total biomass
in plot per ha
 Belowground Tree Carbon
• Use equation based on aboveground tree
biomass (Mokany et al 2006 )
• Use Root-Shoot ratios

AG

Root:Shoot = BG/AG

BG
 Dead Wood Densities
• Calculate dead wood densities for each of
sound, intermediate, rotten samples
collected:
Mass = oven-dried weight (g)
Volume = π x (average diameter/2)2 x average width of
A  r 2
the fresh sample
W2 W1

Density (g/cm3) = Mass (g) / Volume (cm3)

• Calculate average wood densities for each class (S, I, R)
 Estimating Dead Wood Density
Use water displacement method – used for
irregular objects
• Collect samples from all classes:
– Sound
– Intermediate*
– Rotten*
• Place sample in volumetric cylinder
• Record volume of water displaced
by deadwood sample
*It may be difficult to maintain volume of intermediate and rotten samples
stable and impermeable – use plastic wrap.
 Standing Dead Wood Carbon
Standing dead wood carbon
• Class 1 Trees:
– Estimate biomass using allometric equation
– Subtract biomass of leaves:
• 3% of biomass for broadleaf species
• Class 2 Trees:
– To be conservative, estimate biomass of bole only
– Estimate biomass using volume and density:
Volume (m3) = 1/3 *π *h*(r12 + r22 + (r1 *r2))
r1 = radius at base of tree
r2 = radius at top of tree
Biomass = Volume * sound wood density (from samples)
• Multiply mass of sample by scaling factor for the
appropriate nest size
• Convert biomass to carbon
 Lying Dead Wood Carbon
Lying dead wood carbon
• Calculate volume for each density class
separately:

Volume  (m 3 )   2   1
d
2
 d 2
2
     d n
2

 8L 
Where d =diameters of intersecting pieces of
dead wood and L = length of transect
• Calculate carbon of lying dead wood pool
– Biomass in density class (t/ha) =(volume x density)
– Sum biomass of three density classes
– Multiply by carbon fraction to convert to carbon (t/ha)
Herbaceous vegetation,saplings,
Herbaceous vegetation, saplings, & litter
& litter

• Calculate dry mass of sample:

 subsample  dry  mass 

Dry  mass     fresh  mass  of  whole  sample
 subsample  fresh  mass 

• Multiply dry mass of sample by scaling factor

• Convert biomass to carbon
 90% confidence interval
• For normally distributed data, we expect that 95% of data
points will fall within 1.645 standard deviations of the
mean
• Calculate the 95% confidence interval using
– Standard deviation (σ)   
– Sample size (n) 90%CI  1.645 
 n
• Report C stock as mean ± 90%CI
• Uncertainty can also be estimated as a percentage of the
mean (95% CI / mean) x 100  should be <10%
 QA/QC for Data Entry
• Standard Operating Procedures should be
developed and implemented
• Data should be examined for numbers that are
outside range of most numbers as this may be
caused by data entry mistakes
• If problems exist in data entry that cannot be
resolved, this plot should be removed
 QA/QC for Data Storage
• Standard Operating Procedures should be
developed to update and backup all data
• Copies of all data should be stored in a secure
location separate from location of original
data
• Update all electronic data to new types of
data storage as technology changes
Tool to estimate C stocks in all
pools measured using
appropriate regressions and
factors to convert to t C/ha

20
LEAF Technical Training on Forest
Carbon Assessment
Felipe Casarim and Alex Grais
Ecosystem Services Unit, Winrock International
www.winrock.org/ecosystems
carbonservices@winrock.org

Material from this publication may be reproduced provided Winrock

International and the authors are acknowledged as the source.

Casarim, F. and A. Grais. 2013. LEAF Technical Training on Forest Carbon

Assessment.

21