5.5.2 Carbon Emission Mapping
Generally, estimation of carbon emissions in peatlands based on the factor of
emissions is the carbon stock multiplied by the emission factor. Because calculations
based on emission factors with carbon stock obtained from satellite imagery such as
Landsat-8 have the advantages of low cost, rapid speed, and coverage of a wide area,
we developed a method for estimating the carbon emissions that also has the same
advantages. The method uses satellites launched by the National Aeronautics and
Space Administration (NASA) under the name Soil Moisture Active Passive
(SMAP) satellite. Calculations using this method produce a good estimation if the
eddy covariance tower is installed in the area. The measurement shows a strong
correlation between GWL, soil respiration, and soil peat subsidence with carbon
emissions (Hirano et al. 2012; Mezbahuddin et al. 2014). A decrease in GWL of
approximately 0.1 m/year is correlated with carbon emissions of 89 gC/m
2 /year
(Hirano et al. 2012). Peatland carbon emissions occur in the dry season when the
peatlands dry up. The drought level is expressed using soil moisture parameters. A
strong relationship exists between drought and decreasing GWL. Therefore, with the
soil moisture information, we can estimate the intensity of carbon emissions in the
peatlands.
This method is conducting via the following steps:
First (Fig. 5.6a), we classify a peatland into three categories of UF, DF, and DB
(Hamada et al. 2016).
Second (Fig. 5.6b), we take the SMAP image in the form of soil moisture from the
surface to 5 cm depth using GeoTIFF format. The data type used is SMAP Enhanced
L3 Radiometer Global Daily 9 km EASE (Equal Area Scalable Earth) -Grid Soil
Moisture V001 with the boundaries of north 6.46875, south 11.671875, east
142.453125 and west 93.515625. This resolution can be downscaled to 3 km
resolution using the scheme enhanced-resolution SMAP TB products on the 3 km
SMAP project grids (Long et al. 2017). This process is depicted in Fig. 5.6b.
Third (Fig. 5.6c), the GWL map is obtained from the empirical formula between
the GWL and soil moisture with the eight-year time series. Although this relationship was obtained in the Kalimantan peat area, we assume that it applies to all
Indonesian peatlands. The relationship is y UF ¼ 4.89x-1.63, y DF ¼ 4.85x-1.97,
y DB ¼ 2.97x-0.96, where x is soil moisture; y UF , y DF , and y DB are the GWL for
UF, DF, and DB, respectively; and the correlation coefficients are given by 0.67,
0.65, and 0.95, respectively. The result is depicted in Fig. 5.6c.
Fourth (Fig. 5.6d), carbon emission is represented in NEE and is calculated based
on the empirical relationships between the NEEs measured with eddy covariance
instruments for more than 8 years in Central Kalimantan with the GWL parameters
in the same place. The relationship between NEE (φ) and GWL (y) is presented in
equation form as φ UF/DF ¼ À66y-68.74 and φ DB ¼ À420.56y+397.46. This result is
depicted in Fig. 5.6d.
174
N. Tsuji et al.
Generally, estimation of carbon emissions in peatlands based on the factor of
emissions is the carbon stock multiplied by the emission factor. Because calculations
based on emission factors with carbon stock obtained from satellite imagery such as
Landsat-8 have the advantages of low cost, rapid speed, and coverage of a wide area,
we developed a method for estimating the carbon emissions that also has the same
advantages. The method uses satellites launched by the National Aeronautics and
Space Administration (NASA) under the name Soil Moisture Active Passive
(SMAP) satellite. Calculations using this method produce a good estimation if the
eddy covariance tower is installed in the area. The measurement shows a strong
correlation between GWL, soil respiration, and soil peat subsidence with carbon
emissions (Hirano et al. 2012; Mezbahuddin et al. 2014). A decrease in GWL of
approximately 0.1 m/year is correlated with carbon emissions of 89 gC/m
2 /year
(Hirano et al. 2012). Peatland carbon emissions occur in the dry season when the
peatlands dry up. The drought level is expressed using soil moisture parameters. A
strong relationship exists between drought and decreasing GWL. Therefore, with the
soil moisture information, we can estimate the intensity of carbon emissions in the
peatlands.
This method is conducting via the following steps:
First (Fig. 5.6a), we classify a peatland into three categories of UF, DF, and DB
(Hamada et al. 2016).
Second (Fig. 5.6b), we take the SMAP image in the form of soil moisture from the
surface to 5 cm depth using GeoTIFF format. The data type used is SMAP Enhanced
L3 Radiometer Global Daily 9 km EASE (Equal Area Scalable Earth) -Grid Soil
Moisture V001 with the boundaries of north 6.46875, south 11.671875, east
142.453125 and west 93.515625. This resolution can be downscaled to 3 km
resolution using the scheme enhanced-resolution SMAP TB products on the 3 km
SMAP project grids (Long et al. 2017). This process is depicted in Fig. 5.6b.
Third (Fig. 5.6c), the GWL map is obtained from the empirical formula between
the GWL and soil moisture with the eight-year time series. Although this relationship was obtained in the Kalimantan peat area, we assume that it applies to all
Indonesian peatlands. The relationship is y UF ¼ 4.89x-1.63, y DF ¼ 4.85x-1.97,
y DB ¼ 2.97x-0.96, where x is soil moisture; y UF , y DF , and y DB are the GWL for
UF, DF, and DB, respectively; and the correlation coefficients are given by 0.67,
0.65, and 0.95, respectively. The result is depicted in Fig. 5.6c.
Fourth (Fig. 5.6d), carbon emission is represented in NEE and is calculated based
on the empirical relationships between the NEEs measured with eddy covariance
instruments for more than 8 years in Central Kalimantan with the GWL parameters
in the same place. The relationship between NEE (φ) and GWL (y) is presented in
equation form as φ UF/DF ¼ À66y-68.74 and φ DB ¼ À420.56y+397.46. This result is
depicted in Fig. 5.6d.
174
N. Tsuji et al.