QUANTIFYING TROPICAL WETLAND BIOMASS AND CARBON STOCKS
Locating and mapping peatlands:
workshop for national stakeholders
February 1, 2021
https://www2.cifor.org/swamp/
Sustainable Wetlands Adaptation and Mitigation Program (SWAMP)
+ Democratic Republic of the Congo (P);
Ghana (M); Madagascar (M)
Sustainable Wetlands Adaptation and Mitigation Program (SWAMP)
https://www2.cifor.org/global-wetlands/
https://data.cifor.org/dataverse/swamp
Sustainable Wetlands Adaptation and Mitigation Program (SWAMP)
SWAMP Protocol Adopted Internationally:
2013 Supplement to the
IPCC Guidelines for
National GHG
Inventories: Wetlands
Sustainable Wetlands Adaptation and Mitigation Program (SWAMP)
Training on the National Forest Inventory
Field Protocol for Peatlands
January 24-31, 2018 | Mbandaka, DRC
Sustainable Wetlands Adaptation and Mitigation Program (SWAMP)
SWAMP IN SOUTH AMERICA
• Mapping wetlands
and peatlands in the
Andes and the
Amazon
• Developing
degradation maps
• Developing
deforestation maps
• Capacity building
within government
agencies
Sustainable Wetlands Adaptation and Mitigation Program (SWAMP)
CHALLENGES FOR MAPPING
PEATLANDS IN SOUTH AMERICA
• Global methods (e.g
Gumbricht et al. 2017) have
limitations, especially in the
mountains with difficult
access
• Peru has a good ecosystems
map, however, land cover
maps do not include
peatlands, especially in the
Andes
• It is hard to differentiate
peatlands from wetlands
• Because, for example, using
"landsat" they look very similar
• Even looking at them, they are
very similar, until it is examined
belowground
Sustainable Wetlands Adaptation and Mitigation Program (SWAMP)
OUR SOLUTION
Groundtruthing, including peat
sampling
Adding multidate
radar methods
Modelling with
“random forest”
provides supervised
classification
With the map and the soil
samples, we can
estimate C stocks
Image/GIS Layer
Stack
Random Forests
Classifier
SRTM/ASTER DEM
Topographic Position
Index
Radiometric Terrain
Correction
SAR
Multi-Date
Stack
Training/ Validation
Classified Image
Aerial Image
Interpretation
Field Data
40 m x 50 m
plots
Landsat
Multi-Date
Stack
SAR
Multi-Date
Stack
C
HIGHLAND AND
AMAZONIAN WETLAND
MAPS
Andean mapping
> 90% precision
Our map National Ecosystems Map
Peatland area = 108,690 ha Peatland area = 21,549 ha
PERUVIAN AMAZON WETLANDS MAPS
Sustainable Wetlands Adaptation and Mitigation Program (SWAMP)
PERUVIAN PEATLANDS
 Peatlands mostly concentrated in Loreto (Amazon) &
covered by palm swamp forests (PSF)
→ M. flexuosa palms cut for fruit
harvesting & worm collection
→Critical lack of knowledge on:
Extension & distribution of
degradation
Impacts of degradation on C
stocks and GHG emissions
 PSF under degradation threat over past 30 years
Sustainable Wetlands Adaptation and Mitigation Program (SWAMP)
Extent of degradation &
impacts on vegetation C
stocks (Hergoualc’h et al., 2017)
 Shift: palm-dominated to woody
tree-dominated forest
 Biomass reduced considerably
with degradation
-
26
%
0
40
80
120 Biomass C stock
(Mg C ha-1)
Other palms
-
44
%
 73% of PSF area classified as
degraded in 350,000 ha of study
 Highest degradation towards the
north, lowest inside Pacaya Samiria
reserve
→ Iquit
os
Pole forest in the hydromorphic
area have ~1390 Mg C ha-1
Palm swamps have ~828
Mg C ha-1
OUR MAP
User accuracy Draper 2014 This map
Herbaceous-shrub swamp 94.0% 99.0%
Palm swamp forest 78.9% 98.3%
Pole forest 80.0% 95.0%
Floodplain alluvial forest 80.1% 84.9%
Upland forest 99.7% 91.5%
Carbon (Pg) 3.14 ~3.47
Sustainable Wetlands Adaptation and Mitigation Program (SWAMP)
DEFORESTATION MAP
Bourgeau-Chavez, L.L., et al. Unpublished data
Sustainable Wetlands Adaptation and Mitigation Program (SWAMP)
COLLABORATION
• We are working with national institutions within MINAM to
develop maps at a national level, focusing on building
capacity within the institutes.
• INAIGEM for the National Bofedales Inventory, for the Andean region.
• IIAP to create a wetlands and peatlands map for the Amazon region.
Sustainable Wetlands Adaptation and Mitigation Program (SWAMP)
POTENTIAL TO IMPROVE WETLANDS AND
PEATLANDS MAPS AT A NATIONAL LEVEL
• For a national map we
need to:
• Enhance the local
capacity to use these
methods
• Have more groundtruthing
points
• An opportunity to
collaborate with the
National forest inventory
(INFFS)?
• Rapid sample methods or
peat probing.
• Apply them at a national
level
MARCO METODOLÓGICO DEL INVENTARIO
NACIONAL FORESTAL Y DE FAUNA SILVESTRE 2016
Sustainable Wetlands Adaptation and Mitigation Program (SWAMP)
KEY MESSAGES
• There is more peat in the Peruvian Andes and in the Amazonian pole
forest than we previously thought.
• The amount of carbon in peat rivals that of rainforest biomass and
remains an important sponge for atmospheric CO2, which, unlike
forests, can grow for millennia if protected. This natural resource is
also a potential economic resource in the context of payments for
carbon sequestration. But to manage this resource, we need an
accurate map.
• With good field data, and more field points, these maps can be
improved and expanded to the national scale, providing a basis for
good management.
• Deforestation is lower than expected in the hydromorphic area, in
part because of the limited accessibility.
• SWAMP global database and course resolution map publicly available
• CNIAF and DIAF teams trained on SWAMP protocols for data
collection in peatland forests
MERCI
GRACIAS
THANK YOU
Denis Sonwa | D.Sonwa@cgiar.org
Olivia Freeman|olivia.freeman@fs-ip.us
Ana Maria Planas | amplanas.silvacarbon@gmail.com
Sustainable Wetlands Adaptation and Mitigation Program (SWAMP)

Locating and mapping peatlands: workshop for national

  • 1.
    QUANTIFYING TROPICAL WETLANDBIOMASS AND CARBON STOCKS Locating and mapping peatlands: workshop for national stakeholders February 1, 2021 https://www2.cifor.org/swamp/
  • 2.
    Sustainable Wetlands Adaptationand Mitigation Program (SWAMP) + Democratic Republic of the Congo (P); Ghana (M); Madagascar (M)
  • 3.
    Sustainable Wetlands Adaptationand Mitigation Program (SWAMP) https://www2.cifor.org/global-wetlands/ https://data.cifor.org/dataverse/swamp
  • 4.
    Sustainable Wetlands Adaptationand Mitigation Program (SWAMP) SWAMP Protocol Adopted Internationally: 2013 Supplement to the IPCC Guidelines for National GHG Inventories: Wetlands
  • 5.
    Sustainable Wetlands Adaptationand Mitigation Program (SWAMP) Training on the National Forest Inventory Field Protocol for Peatlands January 24-31, 2018 | Mbandaka, DRC
  • 6.
    Sustainable Wetlands Adaptationand Mitigation Program (SWAMP) SWAMP IN SOUTH AMERICA • Mapping wetlands and peatlands in the Andes and the Amazon • Developing degradation maps • Developing deforestation maps • Capacity building within government agencies
  • 7.
    Sustainable Wetlands Adaptationand Mitigation Program (SWAMP) CHALLENGES FOR MAPPING PEATLANDS IN SOUTH AMERICA • Global methods (e.g Gumbricht et al. 2017) have limitations, especially in the mountains with difficult access • Peru has a good ecosystems map, however, land cover maps do not include peatlands, especially in the Andes • It is hard to differentiate peatlands from wetlands • Because, for example, using "landsat" they look very similar • Even looking at them, they are very similar, until it is examined belowground
  • 8.
    Sustainable Wetlands Adaptationand Mitigation Program (SWAMP) OUR SOLUTION Groundtruthing, including peat sampling Adding multidate radar methods Modelling with “random forest” provides supervised classification With the map and the soil samples, we can estimate C stocks Image/GIS Layer Stack Random Forests Classifier SRTM/ASTER DEM Topographic Position Index Radiometric Terrain Correction SAR Multi-Date Stack Training/ Validation Classified Image Aerial Image Interpretation Field Data 40 m x 50 m plots Landsat Multi-Date Stack SAR Multi-Date Stack C
  • 9.
  • 10.
    Andean mapping > 90%precision Our map National Ecosystems Map Peatland area = 108,690 ha Peatland area = 21,549 ha
  • 11.
  • 12.
    Sustainable Wetlands Adaptationand Mitigation Program (SWAMP) PERUVIAN PEATLANDS  Peatlands mostly concentrated in Loreto (Amazon) & covered by palm swamp forests (PSF) → M. flexuosa palms cut for fruit harvesting & worm collection →Critical lack of knowledge on: Extension & distribution of degradation Impacts of degradation on C stocks and GHG emissions  PSF under degradation threat over past 30 years
  • 13.
    Sustainable Wetlands Adaptationand Mitigation Program (SWAMP) Extent of degradation & impacts on vegetation C stocks (Hergoualc’h et al., 2017)  Shift: palm-dominated to woody tree-dominated forest  Biomass reduced considerably with degradation - 26 % 0 40 80 120 Biomass C stock (Mg C ha-1) Other palms - 44 %  73% of PSF area classified as degraded in 350,000 ha of study  Highest degradation towards the north, lowest inside Pacaya Samiria reserve → Iquit os
  • 14.
    Pole forest inthe hydromorphic area have ~1390 Mg C ha-1 Palm swamps have ~828 Mg C ha-1 OUR MAP User accuracy Draper 2014 This map Herbaceous-shrub swamp 94.0% 99.0% Palm swamp forest 78.9% 98.3% Pole forest 80.0% 95.0% Floodplain alluvial forest 80.1% 84.9% Upland forest 99.7% 91.5% Carbon (Pg) 3.14 ~3.47
  • 15.
    Sustainable Wetlands Adaptationand Mitigation Program (SWAMP) DEFORESTATION MAP Bourgeau-Chavez, L.L., et al. Unpublished data
  • 16.
    Sustainable Wetlands Adaptationand Mitigation Program (SWAMP) COLLABORATION • We are working with national institutions within MINAM to develop maps at a national level, focusing on building capacity within the institutes. • INAIGEM for the National Bofedales Inventory, for the Andean region. • IIAP to create a wetlands and peatlands map for the Amazon region.
  • 17.
    Sustainable Wetlands Adaptationand Mitigation Program (SWAMP) POTENTIAL TO IMPROVE WETLANDS AND PEATLANDS MAPS AT A NATIONAL LEVEL • For a national map we need to: • Enhance the local capacity to use these methods • Have more groundtruthing points • An opportunity to collaborate with the National forest inventory (INFFS)? • Rapid sample methods or peat probing. • Apply them at a national level MARCO METODOLÓGICO DEL INVENTARIO NACIONAL FORESTAL Y DE FAUNA SILVESTRE 2016
  • 18.
    Sustainable Wetlands Adaptationand Mitigation Program (SWAMP) KEY MESSAGES • There is more peat in the Peruvian Andes and in the Amazonian pole forest than we previously thought. • The amount of carbon in peat rivals that of rainforest biomass and remains an important sponge for atmospheric CO2, which, unlike forests, can grow for millennia if protected. This natural resource is also a potential economic resource in the context of payments for carbon sequestration. But to manage this resource, we need an accurate map. • With good field data, and more field points, these maps can be improved and expanded to the national scale, providing a basis for good management. • Deforestation is lower than expected in the hydromorphic area, in part because of the limited accessibility. • SWAMP global database and course resolution map publicly available • CNIAF and DIAF teams trained on SWAMP protocols for data collection in peatland forests
  • 19.
    MERCI GRACIAS THANK YOU Denis Sonwa| [email protected] Olivia Freeman|[email protected] Ana Maria Planas | [email protected] Sustainable Wetlands Adaptation and Mitigation Program (SWAMP)