Soil Carbon
Modelling
Budiman Minasny
Alex McBratney
Jose Padarian
The University of Sydney
Page 1
Background – Why the interest in soil carbon?
The University of Sydney
Page 2
Soil carbon
The University of Sydney
Page 3
Why Soil Carbon ‘Gold Rush’?
Soil C pool
Organic C 1550 Pg
Inorganic C 950 Pg
Atmospheric
pool
760Pg
Biotic
pool
610 Pg
1 Pg = 1015 g = 1 billion ton
The University of Sydney
Source: Lal (2006)
Page 4
Paris Climate Change Conference
–  Soil Carbon 4 per mille
8.9
8.9
billion tonne C
Global
CO2 emissions from fossil fuels
The University of Sydney
2400
billion tonne C
Organic carbon stored in the soil globally
(up to 2 m)
2400
= 4‰
Amount of C stock increase
needed to
offset CO2
emission
Page 5
Soil C 4 per mille
Agricultural practices that can sequester soil carbon
The University of Sydney
Page 6
Soil C 4 per mille
The University of Sydney
Page 7
–  The need for better and more cost effective measurement of soil C
–  Auditing SOC
–  Modelling SOC
–  Monitoring SOC
The University of Sydney
Page 8
Empirical models
The University of Sydney
Page 9
Digital soil mapping
Soil observations
C = f(s,c,o,r,p,a,n)
Spatial Soil
Prediction Functions
scorpan layers
Existing Soil
maps
Remote Sensed
Images
DEM
DTM
Climate
The University of Sydney
Organic C: 0-5 cm
Organic C: 5-15 cm
C stock
Organic C: 15-30 cm
Organic C: 30-60 cm
Page 10
Digital soil mapping
Prediction
Prediction variance
40
9
0
0
1.5
0
3 km
C stock
The University of Sydney
Page 11
Sampling Design for Soil C Auditing
De Gruijter et al., 2016. Farm Scale Soil C Auditing. Geoderma.
The University of Sydney
Page 12
Semi Empirical/Mechanistic models
The University of Sydney
Page 13
Global Soil C Assessment
The University of Sydney
Page 14
Global Soil C Assessment
–  Models based on land cover change
–  dC/dt = A – k. C
The University of Sydney
Page 15
The University of Sydney
Page 16
The University of Sydney
Page 17
Process-based models
The University of Sydney
Page 18
Struc-C
•  There are influences and feedbacks between Soil
Organic Carbon and soil structure
(e.g. Edwards and Bremner, 1967; Field, 2000; Kleber
et al, 2007; Six et al., 2004)
•  Soil structure is rarely considered in soil carbon
models
◊ How does SOC influence the soil aggregation?
◊ Building a better structure (model)
The University of Sydney
Page 19
FOM input
Per month
RothC
DPM
CO2
Decay
BIO
RPM
CO2
Decay
HUM
BIO
Decay
IOM
HUM
Coleman and Jenkinson, 1999
CO2
CO2
DPM
Struc-C
FOM input
Per month
CO2
AC 1
CO2
AC 2
AC 3
RPM
CO2
NCOC
CO2
The University of Sydney
Page 20
Aggregation
–  The smallest aggregates are composed of
organo-mineral associations. Strong clay
to SOC bonds
–  The clustering of these smaller aggregates
forms larger aggregates, which in turn
provides physical protection of the SOC
between these aggregates from microbial
attack.
The University of Sydney
Page 21
Forest to plantation
Plantation to cropland
Type 3
Type 2
Type 1
The University of Sydney
Page 22
The University of Sydney
Page 23
Simulation
Cropped Restored prairie
Tallgrass prairie
7
6
5
SOC (%)
4
Type-3
3
Type-2
2
Type-1
1
0
100
200
NCOC
300
400
500
600
Years
The University of Sydney
Page 24
10
1.3
9
1.2
8
Bulk density (g/cm3)
1.1
Total OC (%)
7
6
5
1
0.9
0.8
4
0.7
3
2
0
5
10
15
20
25
Years
30
35
40
45
50
0
5
10
15
20
25
Years
30
35
40
45
Years since restoration
The University of Sydney
Page 25
50
60
4.5
Micro within
macroaggregate C
4
50
Micro within
macroaggregate C
3
30
Macroaggregate C
20
C Agg(%)
Aggregates(%)
40
3.5
2.5
2
Macroaggregate C
1.5
1
10
0.5
Microaggregate C
0
0
5
10
15
The University of Sydney
20
25
Years
30
35
40
45
50
0
Microaggregate C
0
5
10
15
20
25
Years
30
35
40
45
Page 26
50
Other models
–  Carbon, aggregation, and structure turnover (CAST) , Stamati
et al. 2013
AggModel, Segoli et al. 2013
The University of Sydney
Page 27
Struc-C Improvement
–  A better definition of Aggregate pools
–  Reconciling measured & modelled aggregates
–  Distribution with depth
CO2
DPM
FOM input
Per month
CO2
CO2
Micro
Macro
RPM
CO2
NCOC
CO2
The University of Sydney
Page 28
Soil C Landscape model
–  3D Soil genesis model, millennial time scale
–  C -> Simple 2 Compartment model, NPP production,
erosion, vertical mixing, productivity feedback
The University of Sydney
A quantitative model for integrating landscape evolution and soil formation
T. Vanwalleghem, U. Stockmann, B. Minasny, A.B. McBratney. JGR
Page 29
Soil-landscape model
The University of Sydney
Page 30
Conclusions
–  4 per 1000: Soil C is relevant for mitigation of climate change
–  To achieve 4 per mille Soil C initiative, we need to be able to
measure it with confidence, model the C change & monitor it.
–  Soil C & structure still lacks a quantitative model.
–  Opportunities for greater collaboration
The University of Sydney
Page 31