RESEARCH
R E P O R T
Carbon, Land and Water:
101 A Global Analysis of the
Hydrologic Dimensions of Climate
Change Mitigation through
Afforestation/Reforestation
Robert J. Zomer, Antonio Trabucco, Oliver van Straaten
and Deborah A. Bossio
I n t e r n a t i o n a l
Water Management
I n s t i t u t e IWMI is a Future Harvest Center
supported by the CGIAR
Research Reports
IWMI’s mission is to improve water and land resources management for food,
livelihoods and nature. In serving this mission, IWMI concentrates on the integration
of policies, technologies and management systems to achieve workable solutions
to real problems—practical, relevant results in the field of irrigation and water and
land resources.
The publications in this series cover a wide range of subjects—from computer
modeling to experience with water user associations—and vary in content from
directly applicable research to more basic studies, on which applied work ultimately
depends. Some research reports are narrowly focused, analytical and detailed
empirical studies; others are wide-ranging and synthetic overviews of generic
problems.
Although most of the reports are published by IWMI staff and their collaborators,
we welcome contributions from others. Each report is reviewed internally by IWMI’s
own staff and Fellows, and by external reviewers. The reports are published and
distributed both in hard copy and electronically (www.iwmi.org) and where possible
all data and analyses will be available as separate downloadable files. Reports may
be copied freely and cited with due acknowledgment.
Research Report 101
Carbon, Land and Water: A Global Analysis
of the Hydrologic Dimensions of Climate
Change Mitigation through Afforestation/
Reforestation
Robert J. Zomer, Antonio Trabucco, Oliver van Straaten
and Deborah A. Bossio
International Water Management Institute
P O Box 2075, Colombo, Sri Lanka
i
IWMI receives its principal funding from 58 governments, private foundations, and
international and regional organizations known as the Consultative Group on International
Agricultural Research (CGIAR). Support is also given by the Governments of Ghana,
Pakistan, South Africa, Sri Lanka and Thailand.
The authors: Robert J. Zomer is a Senior Researcher, Deborah A. Bossio is Principal
Researcher, and Antonio Trabucco is a Research Associate, all of the International Water
Management Institute (IWMI). Oliver van Straaten is a Research Associate at the World
Agroforestry Center, seconded to IWMI.
Acknowledgements: The authors would like to acknowledge the support of the International
Water Management Institute (IWMI), and the World Agroforestry Center. Local case studies
were completed with research support provided by a grant from the European Union/
EuropeAid (B7-6200/2002/069-203/TPS) within the framework of the ENCOFOR Project
(see http://www.joanneum.at/encofor). We would like to thank the institutions responsible
for leading the work of the ENCOFOR project in the participating countries for facilitating
the data collection: in Bolivia, Fundación Centro Técnico Forestal (CETEFOR), and in
Ecuador, Programa Face de Forestación (PROFAFOR). We would personally like to thank
the ENCOFOR Project team: Igino Emmer and Bart Muys for coordination and project
management; the ENCOFOR country partners in Ecuador, Bolivia, and Uganda: Luis
Fernando, Anko Stilma, Timm Tennigkeit; Lou Verchot of the World Agroforestry Center;
and our colleagues from the Joanneum Institute: Bernhard Schlamadinger, Hannes
Schwaiger, and Neil Bird for their valuable contributions to this work. Many thanks to
Vladimir Smakhtin of IWMI and Bart Muys of K.U. Leuven for valuable advice, reviews
and comments on the manuscript. Much appreciation goes to the IWMI RS/GIS Team -
Sadir Mohideen, Anjitha Senarath, Li Yuanjie, and Aminul Islam. We extend our appreciation
to the CGIAR-Consortium for Spatial Information (CGIAR-CSI; see http://csi.cgiar.org) for
hosting the ENCOFOR Online Analysis Tool.
Zomer, R. J.; Trabucco, A.; van Straaten, O.; Bossio, D. A. 2006. Carbon, land and water:
A global analysis of the hydrologic dimensions of climate change mitigation through
afforestation/reforestation. Colombo, Sri Lanka: International Water Management Institute.
44p. (IWMI Research Report 101)
/ climate change / afforestation / reforestation / land use / water balance / models / trees
/ forests / water use / hydrology / ecosystems /
ISSN 1026-0862
ISBN 92-9090-641-3
ISBN 978-92-9090-641-4
Copyright © 2006, by IWMI. All rights reserved.
Cover photographs by Robert Zomer and Oliver van Straaten. Geospatial analysis of
carbon, land, and water under the Kyoto Protocol’s Clean Development Mechanism
afforestation/reforestation provisions has been combined with fieldwork and case studies
to develop land suitabilty models and provide estimates of hydrologic impacts.
Please send inquiries and comments to: iwmi@cgiar.org
Contents
Summary          v
Introduction          1
Background          2
Research Objectives          5
Methods          6
Results and Discussion          14
Conclusion          31
Literature Cited          33
iii
iii
Acronyms
AI Aridity Index
AR Afforestation/Reforestation
CDM Clean Development Mechanism
CDM-AR Clean Development Mechanism - Afforestation/Reforestation
CER Certified Emissions Reduction
COP Conference of the Parties to the UNFCCC
DEM Digital Elevation Model
FAO Food and Agriculture Organization of the United Nations
GHG Greenhouse Gas
GIS Geographic Information System
GLCF Global Land Cover Facility (University of Maryland)
IPCC Intergovernmental Panel on Climate Change
ISLSCP International Satellite Land-Surface Climatology Project
IUCN World Conservation Union (formerly International Union for the Conservation of Nature and
  Natural Resources)
IWMI International Water Management Institute
KP Kyoto Protocol
LULUCF Land Use, Land Use Change and Forestry
MODIS Moderate Resolution Imaging Spectrometer
NIMA National Imagery and Mapping Agency (USGS)
PCF World Bank Prototype Carbon Fund
SBSTA Subsidiary Body for Scientific and Technological Advice
SRTM Shuttle Radar Topography Mission
UNEP United Nations Environment Programme
UNFCCC United Nations Framework Convention on Climate Change
USGS United States Geological Survey
WDPA World Database of Protected Areas
WHO World Health Organization
Summary
Climate change and global warming have 1,000 meters (m) in elevation and of moderate
become familiar notions throughout the world, as productivity.
the profound impact that human activities have  If converted to forest, large areas deemed
made on global biogeochemical cycles is suitable for CDM-AR would exhibit increases in
increasingly recognized. The global carbon cycle actual evapotranspiration and/or decreases in runoff,
has received much international attention as it i.e., a decrease in water potentially available off-site
has become increasingly obvious that increased for other uses. This is particularly evident in drier
levels of CO2 in the atmosphere are causing areas, the semi-arid tropics, and in conversion from
changes in our climate at an alarming rate. The grasslands and subsistence agriculture. However,
Kyoto Protocol is an international effort aimed at major direct impacts of CDM-AR at the global and
mitigating climate change through the reduction regional scales on water resources and food
of greenhouse gas emissions into the security are ascertained as unlikely, primarily due to
atmosphere. Within the Kyoto Protocol, the the UNFCCC mandated cap on CDM-AR at one
Clean Development Mechanism (CDM) is an percent per annum of total emission obligations.
instrument which is intended to reduce However, significant changes in CDM-AR rules
greenhouse gas emissions, while assisting affecting the number of projects or amount of land
developing countries in achieving sustainable that could eventually be under CDM-AR, should
development, with the multiple goals of poverty take into account these potential impacts on the
reduction, environmental benefits and cost- hydrological cycle, and related food security
effective emission reductions. The CDM allows issues. At the local and project level scale,
for a small percentage of emission reduction impacts on water use was substantial. It was
credits to come from reforestation and evident that CDM-AR projects can benefit from
afforestation (CDM-AR) projects. identifying locally optimal locations for tree
In this report, we articulate the ‘hidden’ water plantations that maximize the positive aspects of
dimensions of international efforts to mitigate increased ‘green water’ vapor flows and reduced
climate change through multilateral treaties runoff.
through a global analysis of land suitability and This report highlights the potentially
water use impacts of CDM-AR carbon ‘sink’ significant impacts on the hydrologic cycle and
projects. Large amounts of land were identified the importance of considering secondary effects,
globally as biophysically suitable and meeting the particularly with regard to water, resulting from the
CDM-AR eligibility criteria. The eco-sociologic widespread adoption of global climate change
characteristics of these suitable areas were mitigation measures. It is recommended that the
examined, with results showing that much of this implicit hydrologic dimensions of climate change
land is under rain-fed and/or subsistence mitigation should be more formally articulated
agriculture or savannah land. Large amounts of within the international environmental conventions,
suitable land exhibited relatively low population and recognized within future UNFCCC
densities. Generally, most of this land is below negotiations on the CDM-AR provisions.
v

Carbon, Land and Water: A Global Analysis of the
Hydrologic Dimensions of Climate Change Mitigation
through Afforestation/Reforestation
Robert J. Zomer, Antonio Trabucco, Oliver van Straaten and Deborah A. Bossio
Introduction
Human activities have profoundly affected global other legally mandated frameworks to minimize
biogeochemical cycles and it is widely predicted and mitigate impacts, including such agreements
that human induced climate change will as the United Nations Framework Convention on
significantly affect the biosphere of our planet. Climate Change (UNFCCC), the Convention on
The global carbon cycle has received the most Climate Change, with the Kyoto Protocol (KP),
attention in recent years as it has become the Convention on Biological Diversity, the
evident that increased levels of CO2 in the Convention to Combat Desertification, and more.
atmosphere are causing changes in our climate at Each sets up institutions and mitigation
an alarming and accelerating rate (IPCC 1996; measures that address global change issues and
IPCC 2001). While many factors play into the processes, and create mechanisms which are
complex equation of the impact of greenhouse legally binding to the signatory countries. These
gas (GHG) emissions on the concentration of institutions and measures have, however, complex
gases in the atmosphere, such as buffering by interactions with real world multi-process, multi-
the world’s oceans, there are two essential scale conditions, and can have both intended and
mitigation strategies available: emission unintended effects on carbon and other
reductions, or fixation of atmospheric CO2 into so- biogeochemical processes, but also on hydrologic
called sinks, mainly biomass and ecosystems cycles. In this report we articulate the implicit
through photosynthesis. When this carbon fixation hydrologic dimensions of international efforts to
is semi-permanent, such as in forests, or mitigate climate change, specifically
recalcitrant soil organic matter, it is termed investigating potential impacts of the Clean
‘carbon sequestration’. Partial solutions to Development Mechanism - Afforestation/
increased atmospheric CO2 concentrations can Reforestation (CDM-AR) provisions of the KP.
therefore be found in sequestering carbon in The CDM-AR allows for carbon sequestration
terrestrial ecosystems (IPCC 2000). Forests and offsets of emission reduction obligations for the
trees are important in this regard because they developed countries, through the purchase of
store large quantities of carbon in vegetation and ‘carbon credits’ from afforestation/reforestation
soils. Forests are both sources of atmospheric projects in developing countries. These
CO2, when disturbed by natural or human causes, activities are generally referred to as ‘sink’
and sinks when vegetation and soil carbon projects. This study delineates the potentially
accumulate after afforestation or natural suitable areas for CDM-AR projects globally,
revegetation. describes the socio-ecological characteristics of
International efforts have mobilized to these suitable lands, and estimates the impacts
address climate change and other global of CDM-AR on global, regional and local water
environmental problems with global treaties and cycles.
1
Background
In 1992, the United Nations Framework force, and the first commitment period is from
Convention on Climate Change (UNFCCC) was 2007-2012, much effort has already gone into
the first international convention to recognize the developing CDM and CDM-AR projects. Funds
problem of climate change. It set out the have been set up to support CDM projects around
objective of stabilizing GHG concentrations in the the world, such as the World Bank Prototype
atmosphere to prevent dangerous interference Carbon Fund (PCF) and the BioCarbon Fund,
with climate. The risks of climate change to food more specifically for CDM-AR. In addition, there
production and the importance of adaptation were have been various capacity building activities for
particularly highlighted. The UNFCCC primarily recipient countries and substantial private sector
encouraged developed countries to stabilize activity has developed (Huq 2002).
emissions. In 1997, specific legally-binding
targets and timetables for cutting emissions were
developed and adopted as part of the KP to the Clean Development Mechanism
Convention (UNFCCC). The KP allows for various
mechanisms to achieve these targets, including One of the main purposes of the CDM is to
the Clean Development Mechanism (CDM). CDM assist developing countries in achieving
projects provide credit for financing emissions- sustainable development, with the multiple goals
reducing or emissions-avoiding projects in of poverty reduction, environmental benefits and
developing countries. It is hoped that the CDM cost-effective emissions reductions. The CDM is
will be an important new avenue through which intended to provide a market vehicle through
governments and private corporations can which developed countries with high rates of CO2
promote sustainable development and transfer of emissions (referred to as Annex I Countries) can
clean technologies. Land use, land use change, offset part of their emissions by purchasing
and forestry (LULUCF) activities were included in carbon credits in developing countries. Bioenergy
the KP CDM instrument, recognizing the role of production is one CDM strategy in which biomass
land use, and particularly forests, in regulating is grown (CO2 is fixed) and then used for energy
carbon cycles (Brown et al. 2002). The ability of production (CO2 is released again), thus they
forests (and land) to be both a source and sink substitute CO2 neutral energy for fossil fuel
for carbon allow for manipulation of these energy. CDM sink projects, unlike bioenergy or
processes through forest management and other clean technology transfer projects, require that
human activities, at a significant scale, i.e., carbon be sequestered into semi-permanent
meaningful in terms of climate change mitigation. ‘sinks’, primarily by growing trees, that is,
However, the inclusion of these so-called ‘sink currently through afforestation and reforestation
projects’ and the rules governing eligibility of (CDM-AR) projects. There is considerable
LULUCF carbon offset credits were, and are, optimism in developing countries and the
controversial, producing ample debate during the development community that the potential
various rounds of negotiations (Kolshus 2001; investments represented by CDM sink projects
Kolshus et al. 2001; Forner and Jotzo 2002; Jung can be a boon for rural development and
2003). Concerns center on whether CDM is a environmental protection, if properly directed and
(too) cheap or easy way for Annex I Countries to monitored. Many countries are already heavily
avoid actual emission reductions, and that CDM- involved in planning or implementing pilot projects
AR has a higher risk of leakage and and numerous research programs are underway to
unsustainable practice (Greenpeace 2003). understand and delineate how best to implement
Although the KP has only recently entered into CDM-AR (see http://www.joanneum.at/encofor).
2
Possible afforestation/reforestation activities Environmental and Social Issues of
fall into the following CDM-eligible categories: CDM-AR
• New, large-scale, industrial plantation
Reforestation and/or afforestation represents a
• Introduction of trees into existing agricultural fundamental change in the local ecological
systems (agroforestry) landscape and can have unintended
• Small-scale plantations by landowners consequences or contribute to ecosystem
• Establishment of woodlots on communal degradation. Loss of biodiversity, or other
lands ecosystem services, can result from
establishment of extensive fast growing plantation
• Rehabilitation of degraded areas through tree forests that are economically favored in terms of
planting or assisted natural regeneration low costs per return in fixed carbon. Additionally,
• Reforestation of marginal areas with native some activities may increase erosion, through
species (e.g., riverine areas, steep slopes, disturbances caused by planting, establishment,
around and between existing forest fragments and building of access roads.
through planting and natural regeneration) CDM-AR projects can also have negative
• Establishment of biomass plantation for impacts on rural societies and local economies
energy production and the substitution of where people are dependent upon project area
fossil fuels resources. For example, indigenous land claims
may be infringed when treaties and agreements
Related forestry activities not eligible under are signed at the national level without taking into
the CDM include forest conservation, improved account local institutions or how benefits might
forest management, reduced impact logging, and be equitably shared. Changes in local economic
enrichment planting. Only afforestation/ activity can also affect key factors in sustainable
deforestation is accepted as eligible, as agreed at development such as gender workloads (for
COP 7 in Marrakech (UNFCCC 2002a; UNFCCC example, increasing women’s workload by forcing
2002b). them to go further for firewood and water).
Sink projects continue to be controversial and Projects must engage local population in finding
developing the rules governing their inclusion into alternative sources of livelihood, if these are
global climate change treaties has been long and affected, or provide adequate compensation
arduous. Compared to the CDM technology (Smith and Scherr 2002). Effective carbon sink
transfer activities, CDM-AR projects involve a projects must be integrated into local sustainable
fundamental change in land use. Technology development, and involve far more than simply
transfer makes an activity more efficient and/or planting trees, including concern for off-site
less dependent on non-renewable energy sources. impacts on resources.
Reforestation and/or afforestation is fundamentally In response to these concerns and other
different, implying the cessation of one land use potential negative aspects associated with CDM-
activity and its substitution with another, thus AR, several organizations have highlighted
presenting several unique challenges in both important social justice and environmental
carbon accounting and implementation. To make conservation aspects that are to be evaluated
CDM-AR a positive development vehicle, rules early in project cycles (see http://www.climate-
were agreed upon and methodologies are being standards.org). One such environmental and
developed that attempt to reduce the risk of social issue that has thus far been generally
‘perverse incentives’ that may result in social or overlooked is the water use dimension of carbon
environmental harm, and that adequately verify sequestration projects. Most terrestrial carbon
carbon sequestration, local environmental and fixation is the result of plant growth and
sustainable development benefits, and secure photosynthesis.  This process requires water from
carbon credits. the ecosystem, which, if an increase in carbon
3
stock is achieved, almost certainly means an Other ecosystem service demands for water,
increase in vapor flows, actual evapotranspiration e.g., increased on-site vapor flows associated
(AET), and local in situ water use. global climate change mitigation, are as yet rarely
considered in these discussions. This is partly
due to an under-appreciation that carbon fixation
Water Supply and Carbon through biomass production will require
Sequestration consumption of water that will then not be
available for other uses. A historical hydrological
Water supply and scarcity has received bias in water accounting considered only surface
increasing attention over the last decade, runoff and groundwater as available water supply
primarily driven by alarming WHO figures (2006) and viewed terrestrial ecosystems and forests as
that 1.1 billion people lack access to safe and water-provisioning rather than water consumptive
affordable water for their domestic use. Many of (Falkenmark and Lannerstad 2004). The ongoing
these are the rural poor who lack water not only ‘debate’ on ‘forests and water’ has lately been the
for domestic purposes, but also to sustain subject of much interest and research (CIFOR
agricultural livelihoods (Rijsberman et al. 2006). and FAO 2005), most notably through ecosystem
Numerous projections with regard to water supply evapotranspiration studies (L’vovich and White
and scarcity focus on the rising population and 1990; Gordon et al. 2005), the introduction of the
their needs for domestic and agricultural water. It concepts of green and blue water management in
is estimated, for example, that water diversions agriculture by Falkenmark (1995), Rockstrom et
for agriculture must rise between 12 and 27 al. (1999), and in the forestry sector by Calder
percent by 2025 to meet growing food needs (2000). Only recently have a few studies
(IWMI 2000; FAO 2001b, 2003a, 2003b; highlighted the implications of global climate
Shiklomanov 1998). Many estimates agree that change mitigation strategies on water use
up to two-thirds of the world population will be (Aylward et al. 1998; Calder 2000; Berndes 2002;
affected by water scarcity over the next several Heuvelmans et al. 2005). An analysis of bioenergy
decades (Shiklomanov 1991; Raskin et al. 1997; production concluded that large-scale expansion of
Seckler et al. 1998; Alcamo et al. 1997, 2000; energy crop production would require water
Vorosmarty et al. 2000; Wallace 2000; Wallace consumption equal to that which is currently used
and Gregory 2002). for all crop production (Berndes 2002) and brought
Increasing demands for water to meet direct the implications of this ‘green water’ vapor flow
human needs will be felt most strongly where demand for water into sharp focus.
aquatic and terrestrial ecosystems alike already
suffer from diversions of water for food
production. The conflict between water diversions Forests and Water
to agriculture and maintaining aquatic ecosystems
has received the most attention. Environmental It is generally accepted that tree removal by
flow requirements (Smakhtin et al. 2004) are logging, forest fire, or wind damage increases
increasingly being taken into account to manage runoff (Bosch and Hewlett 1982). Jackson et al.
water allocations, to allow for the perpetuation of (2005) found that plantations decreased stream
natural areas, wildlife and endangered species flow by 227 millimeters (mm) per year globally (52
habitats, and environmentally sensitive wetlands. percent), with 13 percent of streams drying
Links are now also being made between water completely for at least one year. The magnitude
for agricultural food production and water for of this water decrease is proportional to the
terrestrial ecosystem services (Rockstrom et al. percentage of vegetation cover and is due to an
1999). increase in AET, an increase in the net additions
4
to evaporation from interception losses, and an afforestation is not to be necessarily looked at as
increase in the root exploring zone from which a burden for the global hydrological cycle. On-site
water is extracted under trees (Dingman 1993). A hydrological effects of afforestation are mainly
review of catchment experiments (Bosch and positive (reduced runoff and erosion, improved
Hewlett 1982) found that pine and eucalypt microclimate and increased control over nutrient
plantations cause a 40 mm decrease in runoff for fluxes); the off-site effects may be mainly
any 10 percent increase of forest cover with negative (lower base flow), but in many cases
respect to grassland. The equivalent response of these off-site effects of increased in situ vapor
deciduous hardwood and shrubs is 25 and 10 mm flows may be beneficial for downstream users.
decrease in runoff, respectively. Transpiration Gedney et al. 2006 speculate that increases over
from trees can be higher than from shorter the last several decades in total discharge of the
vegetation because tree root systems exploit world’s river systems is a consequence of
deep soil water (Maidment 1992) available during increased CO2 in the atmosphere, which makes
prolonged dry seasons (IPCC 2000). plants more water efficient, although deforestation
Recent references (Gedney et al. 2006; may have played an important part in this
Matthews 2006) support the thesis that phenomena.
Research Objectives
In this research report, we analyzed land and conversion of suitable lands to afforestation/
water use implications of CDM-AR at two scales, reforestation activities.
global and local. Land suitability for CDM-AR was
Specific Objectives:
modeled, as per the existing rules of the first
commitment period, and a simple water balance 1. To delineate areas suitable for CDM-AR,
approach is used to estimate impacts on globally.
hydrological cycles resulting from a change to 2. To characterize suitable areas in both
forestry activities. In addition, socio-ecological biophysical and socio-ecological terms.
characteristics of these suitable areas are
described, including the land use types that 3. To estimate potential impacts of adoption of
currently exist on these lands, and their CDM-AR on global to regional hydrologic
population and ecosystem characteristics. A GIS cycles.
spatial modeling environment is used to delineate 4. To estimate potential impacts of adoption of
biophysical conditions, identify suitable areas for CDM-AR on local hydrologic cycles based on
CDM-AR, and predict hydrologic changes with four in-depth case studies.
5
Methods
The suitability of CDM-AR projects, as per the • Areas classified as irrigated or under other
current proposed guidelines for their application in intensive agricultural production, assuming
developing countries (i.e., Non-Annex I that these areas are already in high value
Countries), is constrained by the current UNFCCC production or their conversion may impact on
guidelines for CDM-AR projects within the first food security
commitment period (2008-2012), the definitions
In addition, areas that are ineligible for CDM-
adopted for forest and forestry activities by
AR due to UNFCCC rules have been excluded
individual countries, and a complex of biophysical
from the analysis:
and socio-economic factors necessary for a
sustainable, socially equitable, and economically • Currently forested areas. A threshold of 30
viable tree growing enterprise. Two main factors percent canopy cover was used as the forest
are reconciled in our analysis: definition, as per results of an earlier analysis
of forest definitions on areas available at a
1. The need to conform to the specific
national scale (Verchot et al. 2006).
guidelines and regulations of the UNFCCC
(e.g., the definition of forest, but also Recently deforested areas, in this case,
explicitly articulated concerns about food areas that are identified as forest in the USGS
security, sustainability and environmental 1993 land use classification but currently exhibit
conservation). a crown cover of less than 30 percent, as per
2. Suitability of the biophysical environment to guidelines that exclude recently deforested areas
support relatively robust biomass production from being eligible for CDM-AR, were delineated
(i.e., fixation of GHG) to make the projects and quantified.
viable and economically feasible. The results of the land suitability analysis are
mapped and tabulated on a national, regional
(sub-continental), and global basis. Results of
Land Suitability Analysis area estimates are articulated by:
• Land Use Types
A spatial modeling procedure was developed and • Population Density
implemented in ArcGIS (ESRI Inc.) using ArcAML
programming language, and used to identify areas • Elevation Zone
meeting a range of suitability criteria as outlined • Aridity Index
below. All areas that are not likely to be suitable
for these projects, due to the following • Net Primary Productivity Class (NPP)
environmental and social factors, have been Environmental and other global geospatial
excluded a priori from our analysis: datasets used within the global analysis include:
• Arid/semi-arid areas with high Aridity Index
(AI < 0.65) (Spatial resolution: 500 m – 1 kilometer (km) / 15
- 30 arc-seconds)
• High elevation areas, above 3,500 m and/or
timberline • VMAP 1 - Country Boundaries (National
Imagery and Mapping Agency) (NIMA 1997)
• Areas covered by water bodies
• Global Ecosystem Land Cover
• Urban areas Characterization Database v. 2.0 (USGS
• Areas classified as various types of tundra 1993)
6
• MODIS Vegetation Continuous Field – Tree Accords’ agreed to at COP 7. Reforestation projects
Cover (Hansen et al. 2003) are allowed only in sites that were not forested on
December 31, 1989 (afforestation generally refers to
• Topography – SRTM DEM (USGS 2004)
sites that have not had forest cover for more than
• World Database on Protected Areas 50 years). The MODIS Vegetation Continuous
(IUCN/UNEP - WDPA Consortium 2004) Fields dataset (Hansen et al. 2003), a global
dataset of tree canopy cover extracted from multi-
• WorldClim (Hijmans et al. 2004)
temporal sequences of MODIS data (year 2001;
• Maximum Available Soil Water resolution 15 arc-seconds) was used in this study
(Digital Soil Map of the World - FAO 1995) to determine currently forested areas. This was
compared with the Land Characteristics Database
• Climate Station Dataset
(USGS 1993) to ascertain recently deforested
(FAOCLIM - FAO 2001a)
areas.
• Gridded Population of the World (2000)
(GPWv3 - CIESIN and CIAT 2005) Elevation limits for CDM projects
• Global Map of Ecosystem Rooting Depth Areas above and approaching timberline were not
(ISLSCP – Schenk and Jackson 2002) considered suitable and were estimated as areas
• MOD17A3 – MODIS Net Annual Primary with average temperature in the growing season
Production (Running et al. 2000) below 6.5o C, according to Korner and Paulsen
(2004) and using length of the growing season
All datasets used for the analyses have been calculations based on the WorldClim dataset
re-projected and processed in two coordinate (Hijmans et al. 2004). Although treeline can
systems, sinusoidal and geographic. The surpass 4,000 meters in certain parts of the
geographic coordinate system preserves landform world, CDM projects have been considered
shapes with a perspective that is generally easily unrealistic at elevations above 3,500 meters.
recognizable to human perception and is therefore Thus, all land above 3,500 meters, (estimated
used for map presentation. The dataset in based on the SRTM DEM) was excluded.
sinusoidal projection was used to calculate zonal
Net Primary Productivity
statistics and carry out areal computations,
because it represents area extent accurately for The MODIS/Terra Annual Net Primary Production
all pixels across latitudes (equal-area projection) dataset (MOD17A3) was obtained from the USGS
while the geographic does not. The cell size for Eros Data Center. MOD17A3 Total Gross Primary
analyses in geographic projection is equal to Productivity is computed using the amount of
0.004497 degrees (15 arc-seconds, ~ 1 km at photosynthetically active radiation (PAR)
equator and 500 m at 60 degrees latitude), while measured by the MODIS instrument. Heinsch et
the cell size for analyses in sinusoidal projection al. (2005) have shown good correlation
(r2
is 500 m.  = 0.859 +- 0.173) between NPP estimated by
MOD17A3 and 38 site years of NPP
Forest Definition, Canopy Cover Percentage, measurements.  Other studies have demonstrated
and Recently Deforested Areas the absence of systematic under- or over-
estimation across different biomes compared to
CDM-AR projects are only eligible and allowed in field observed NPP (Zhao et al. 2005; Turner et
currently non-forested areas. ‘Forests’ are al. 2003). In our analysis, annual NPP grids over
individually defined by each Non-Annex I Country the 2000-2004 period have been aggregated into
as areas within a range of 10-30 percent canopy one average annual NPP dataset and used to
cover, along with a minimum size and height criteria analyze the current productivity of land deemed
(Verchot et al. 2006), based upon the ‘Marrakech suitable for CDM-AR.
7
Water Balance  Model proposed CDM-AR scenarios, and the results are
aggregated into yearly figures.
A spatially distributed Thornthwaite-Mather water
A soil water balance budget is computed as
balance approach (Thornthwaite 1948;
height of water in mm for each month (m), as:
Thornthwaite and Mather 1955) was used to
examine hydrological differences in AET, soil ∆SWCm=EPrecm-AETm-Rm   mm/month         [1]
water content and runoff. This model uses the where: ∆SWCm is the change in soil water content,
average spatially distributed values of monthly
EPrecm is the effective precipitation, AETm is the
precipitation and monthly potential
evapotranspiration (PET), land use classes, soil actual evapotranspiration, and Rm is the runoff
depth and soil water holding capacity, and returns component, which includes both surface runoff and
monthly spatially-distributed raster data subsurface drainage. SWC can never exceed a
representing actual evapotranspiration (AET), maximum value, SWCmax, which is the total SWC
surface runoff (R) and soil water content (SWC). available for evapotranspiration (ET).
All the results are computed on a monthly basis Therefore, the SWC at the end of the month,
f
throughout a year for existing land use and SWC  m   is equal to:
[2]
b b
Where: S  W  C     is the soil water content at the the beginning of the following month, SWC
m m+I
beginning of the month. The SWC at the end All the water exceeding SWCmax is accounted
f
of the month, S  W   C  m   , is set as the SWC at as runoff:
[3]
Monthly Potential Evapotranspiration (PET) were tested to verify which equation performed
Potential evapotranspiration (PET) was estimated the best for the objectives of this analysis:
on a global scale to calculate the Aridity Index Thornthwaite (Thornthwaite 1948), Thornthwaite
(AI) for the land suitability analysis and later used modified by Holland (Holland 1978), Hargreaves
to explore hydrologic impact. PET is a measure (Hargreaves et al. 1985), Hargreaves modified by
of the ability of the atmosphere to remove water Droogers (Droogers and Allen 2002), and the FAO
through ET processes. The FAO introduced a Global Penman-Monteith Dataset (Allen et al.
definition of PET as the ET of a reference crop in 1998). Values of PET estimated using each of
optimal conditions having the following the above five methods were compared to
characteristics: well watered grass with an Penman-Monteith PET values estimated at
assumed height of 12 centimeters (cm), a fixed climate stations in South America and Africa (n =
surface resistance of 70 seconds per meter (s/m) 2288). Based on the results of the comparative
and an albedo of 0.23 (Allen et al. 1998). Five validation for South America (figure 1) and Africa
different methods of calculating PET (table 1) (figure 2), the Hargreaves model was chosen to
8
TABLE 1.
Five different methods of calculating PET were tested to verify which performed the best for the objectives of this analysis:
Thornthwaite (Thornthwaite 1948), Thornthwaite modified by Holland (Holland 1978), Hargreaves (Hargreaves et al.
1985), Hargreaves modified by Droogers (Droogers and Allen 2002), and the FAO Global Penman-Monteith Dataset
(Allen et al. 1998). Results are given as the mean difference (Diff) between observed and predicted estimates, and their
standard deviations (SD).
Comparison of 5 Methods for Estimating PET:
Mean Difference (mm) and Standard Deviation (mm) between Observed and Predicted Values
Holland Thornthwaite Hargreaves Modified Penman-
(Thornthwaite) Hargreaves Montieth FAO
Region Month Diff (SD) Diff (SD) Diff (SD) Diff (SD) Diff (SD)
Africa Jan 71.8 (40.2) 41.6 (33.3) 22.3 (16.1) 24.8 (20.1) 11.1 (12.6)
July 84.4 (41.7) 32.1 (23.7) 20.0 (19.3) 21.1 (19.3) 12.7 (16.0)
South America Jan 69.9 (43.6) 50.5 (32.9) 38.2 (19.2) 41.6 (26.0) 34.9 (26.7)
July 67.3 (35.9) 37.2 (24.7) 27.2 (14.0) 30.4 (20.1) 24.3 (15.1)
Resolution 1 km 1 km 1 km 1 km 20 km
Data Requirements
Average Average Average Average Large Collection
Temperature Temperature Temperature Temperature of Climate Data
Average Average
Extraterrestrial Extraterrestrial
Radiation Radiation
Average Average
Temperature Temperature
Range Range
Average
Precipitation
model PET globally for this study. This method 1 km; resolution of the FAO Penman-Montieth
performed almost as well as the FAO Penman- dataset is 20 km). Hargreaves (1994) uses mean
Monteith, but required less parameterization, and monthly temperature (Tmean), mean monthly
with significantly reduced sensitivity to error in temperature range (TD) and extraterrestrial
climatic inputs (Hargreaves and Allen 2003). This radiation (RA, radiation on top of atmosphere) to
allowed for its application at a finer resolution (at calculate PET, as shown below:
PET   =     0.0023 · RA · (Tmean + 17.8) · TD0.5   (mm/d) [4]
Aridity Index
Usually aridity is expressed as a function of where:
Precipitation, PET, and Temperature (T). In a MAP = mean annual precipitation
classification of climatic zones proposed by the MAE = mean annual evapotranspiration.
UNEP (1997), Aridity Index (AI) is used to
quantify precipitation deficit over atmospheric Monthly values for precipitation, and
water demand: minimum, maximum, and mean temperature
were obtained from the WORLDClim dataset
Aridity Index (AI) = MAP / MAE               [5] (Hijmans et al. 2004) for years 1960-1990, at a
9
FIGURE 1.
Comparison of five methods of calculating PET for South America during two seasons.
10
resolution of 30 arc-seconds, or ~1 km at The maximum amount of soil water available
equator. for ET processes within the plant rooting depth
The global AI dataset produced in the zone, here defined as SWCmax, is equal to the
analysis was compared to the USGS Land SWC at field capacity (SWCfc ) minus the SWC at
Characteristics Database (USGS 1993), and the wilting point (SWCwp ) times the rooting depth.
MODIS Tree Cover Percentage (Hansen et al.
2003) estimates, to obtain an AI threshold. SWCmax = RD * (SWCfc – SWCwp )               [7]
Optimal bioclimatic zones for CDM-AR were
ascertained as AI > 0.65. This lower threshold for where:
suitability represents the moisture range of the
SWC
semi-arid zones (UNEP 1997), which can support max = maximum soil water content available
for ET (mm)
rain-fed agriculture with more or less sustained
RD = rooting depth (mm)
levels of production.
SWC wp = soil water content at wilting point
(mm/mm)
Actual Evapotranspiration and Green Water SWC fc = soil water content at field capacity
Vapor Flows (mm/mm)
Actual evapotranspiration (AET) is the quantity of Soil water content at field capacity and wilting
water that is removed from the soil due to point are available from literature for the various
evaporation and transpiration processes soil texture typologies (Jensen et al. 1990).
(Maidment 1992). AET is dependent on Rooting depth values for the various land use
vegetation characteristics, quantity of water types were modeled by combining rooting depth
available in the soil and soil hydrological of specific vegetation types under irrigated and
properties (mainly soil water retention curves) non-water stress conditions, and their estimated
(Allen et al. 1998): occurrence within those land use types. Rooting
depth of vegetation is likely to be deeper under
AETm = Kveg * Ksoil * PETm    mm/month        [6] water stressed conditions, as water is stored
more in depth in the soil during dry seasons.
where: Rooting depths values for vegetation types under
K irrigated and non-water stress conditions are
soil = reduction factor dependent on volumetric
soil moisture content (0-1) available from the literature (Allen et al. 1998). A
global dataset of ecosystem rooting depth
K (Schenk and Jackson 2002) was used to scale
veg = vegetation coefficient dependent on
vegetation characteristics (0.3-1.3) rooting depth of the various vegetation types to
more realistic water stressed conditions.
The vegetation coefficient (Kveg) is used to The soil stress coefficient (Ksoil) represents
‘correct’ the reference PET for different crops or the ET reduction factor resulting from the limit
vegetation types. Kveg values for the various land imposed by the absolute volumetric soil moisture
use types were modeled by combining Kveg content. The model uses a simple linear soil
coefficients for vegetation types taken from the moisture stress function that is considered
literature, and their estimated occurrence within appropriate for monthly computation (Dyck 1983):
each land use type. Kveg values are available from
literature for agronomic crops (Allen et al. 1998) Ksoil  = SWCm / SWCmax                          [8]
m
and for other vegetation types from various
sources (Allen et al. 1998; Costello and Jones SWCm = soil water content averaged over the
2000; U. S. Bureau of Reclamation 2005). month
11
FIGURE 2.
Comparison of five methods of calculating PET for Africa during two seasons.
12
Effective Precipitation calculated as the gross precipitation (GPrec)
Rain interception is the process by which minus the precipitation intercepted by canopy
precipitation is intercepted by the vegetation cover and litter (Int). The quantity of rain
canopy (canopy interception losses) and litter intercepted is proportional to the interception
(litter interception losses), where it is subject to coefficient Kint, specific for different types of land
evaporation. Interception has an important role in use types, calculated as a fraction of GPrec.
the water budget, as it reduces the amount of There is a wide availability of such coefficients
precipitation available for soil moisture. from literature for different vegetation types (Tate
Additionally, it protects the soil surface from 1996).
erosion by reducing the rainfall energy (Tate For each month EPrecm is calculated as:
1996). The losses due to interception depend on
vegetation type, vegetation cover and the EPrecm = GPrec – Int                            [9]
intensity, duration, frequency and form of
precipitation (Dingman 1993). Observations where: Int is equal to:
derived from several experiments demonstrate
that vegetation interception is a purely mechanic Int = (GPrec * Kint )                               [10]
function of the storage space of vegetation
structure (Wilm 1957). Forests with dense crowns Therefore:
and large leaf areas are expected to have higher
interception losses (IPCC 2000). Interception EPrecm=GPrec–(GPrec*Kint )=GPrec*(1–Kint )   [11]
losses are on average greater for evergreen forest
compared to seasonally leaf-shedding (Schulze We combine the AET and Int components of
1982; Tate 1996) and for fast-growing trees the model to quantify ‘green water’ vapor flows,
compared to slow-growing trees (IPCC 2000). i.e., that portion of precipitation that evaporates
Thin or sparse vegetation shows low values of into the atmosphere, and is not available as
interception (Wilm 1957). Interception values for runoff (or ‘blue water’).
the various land use types were modeled by
combining interception values from the literature Local Water Use Impact
for the various vegetation types (Hamilton and
Rowe 1949; Young et al. 1984; Thurow et al. In order to investigate local and project level
1987; Farrington and Bartle 1991; Calder 1992; Le water use, a similar water balance approach was
Maitre et al. 1999; Schroth et al. 1999), and the applied in four case study sites identified for
estimated occurrence of that specific vegetation CDM-AR (Zomer et al. 2004). These sites
within a land use type. represent a range of biophysical conditions and
Effective precipitation (EPrec), that part of project scenarios, with two sites in Ecuador and
precipitation that adds moisture to the soil, is two in Bolivia (table 2):
TABLE 2.
Socio-ecological characteristics and project scenarios for the four case study sites.
Project Site Ecological Zone Elev Precip Temp Pop Project Type
(m) (mm/yr) (oC)
Tunari NP, Bolivia Sierra 2,800-5,100 900 7-18 22,000 Ecological Restoration
Chapare, Bolivia Amazon 200-1,000 3,000 23-26 9,000 Small Farm Agroforestry
Guamote, Ecuador Sierra 2,900-3,700 700 7-12 5,300 Community Plantations
Coastal Ecuador Tropical Coastal 0-500 1,300 23-25 8,900 Mixed Species Agroforestry
Note: Elev=Elevation; Precip=Precipitation; Temp=Temperature; Pop=Population.
13
Case Studies: proposed CDM-AR scenario, at a resolution of
1. Tunari National Park, Bolivia - Sierras, 30 m for the four case study sites, using both
ecological restoration, native species global and locally available data, and
comparing the proposed CDM-AR project
2. Chapare, Bolivia - Amazon, small farmers’ scenario for the site with the current land use.
agroforestry Tree canopy cover, current and historical, was
3. Guamote, Ecuador - Sierras, community estimated from Landsat TM imagery, and
based plantation scheme elevation was derived from SRTM 90 m DEM
data (available from CGIAR-CSI: http://
4. Coastal, Ecuador - Tropical Coastal Zone, srtm.cgiar.csi.org). Growth characteristics for
mixed/native species agroforestry specific species were obtained from literature
Changes in water cycles were modeled as a and expert knowledge, where available
consequence of land use change to a specific (Zomer et al. 2006).
Results and Discussion
Lands suitable for CDM-AR Scholes 2001; Vrolijk and Grubb 2001; see Jung
2005 for an extensive listing by country). In these
CDM-AR projects are subject to a complex set of global studies, the area available for tree
eligibility guidelines as defined within the plantations is variably estimated at 345 Mha
UNFCCC in order to be certified to provide carbon (Nilsson and Schopfhauser 1995), 465 Mha (Sedjo
emission reduction credits under the CDM. Our and Solomon 1989), and 510 Mha (Nordhaus
global spatial analysis identified all land surface 1991). Nilsson and Schopfhauser (1995) and
areas that meet a minimal set of eligibility Trexler and Haugen (1995) were designated by
criteria, both statutory and biophysical (figure 3). the IPCC Second Assessment Report (Brown et
Results were calculated for the entire world, and al. 1996) as suitable studies for global analysis of
altogether, globally more than 760 million hectares the mitigation potential of forests, including
(Mha) of land were found to be suitable, afforestation/reforestation. The two studies
representing just over nine percent of total land together suggest that 700 Mha of land could be
surface area within the Non-Annex I (developing) available for carbon sequestration and
Countries. Global totals in this paper are reported conservation, globally, including 138 Mha for
as the sum of five regions, which cover most of slowed tropical deforestation, 217 Mha for
the developing (i.e., Non-Annex I) countries with regeneration of tropical forests, and 345 Mha for
significant CDM-AR potential, with the exception plantations and agroforestry. However, Sathaye
of some areas in Central America. Within these and Ravindranath (1998) suggest that 300 Mha
five regions, 725 Mha of land was initially may be available for mitigation in ten tropical and
identified as biophysically suitable. These results temperate countries in Asia, including 181 Mha of
compare well with earlier studies that have asked degraded land for plantation forestry, and 79 Mha
the question how much land is available for of degraded forestland for regeneration. In our
reforestation (Winjum et al. 1998; Nilsson and study, large tracts of suitable land are found in
Schopfhauser 1995; Trexler and Haugen 1995) South America (46 percent of all the suitable
and what is the potential carbon sequestration areas globally) and Sub-Saharan Africa (27
(Yamagata and Alexandrov 2001; Noble and percent), reflecting the greater landmass of these
14
FIGURE 3.
Global map of CDM-AR suitable land within Non-Annex I Countries, as delineated by the land suitability analysis. A
30% crown cover density threshold was used to define forest, and protected areas are not included.
regions, and to a certain extent, lower population as land opportunity costs, access to markets,
densities. Much smaller amounts of land are tenure, or national level infrastructure and
available in Asia, the three Asian regions together support. It is estimated that a substantially
comprising about 200 Mha, compared to more smaller proportion of this identified area will meet
that 330 Mha in South America and almost 200 the more specific criteria which are required to
Mha in Africa. Within the respective regions, the make CDM-AR a viable option for landowners,
amount of available land ranged from only 8 land managers, communities, and/or national
percent of the total land surface area in planners.
Southeast Asia, to more than 19 percent of South
America. Current land use, population and ecosystem
As our suitability estimates are based characteristics of CDM eligible lands
exclusively on biophysical suitability combined To understand the socio-economic and ecological
with UNFCCC requirements, they naturally nature and characteristics of the areas identified
represent an over-estimation of actual areas as suitable, and to better judge the likelihood of
available. Areas that might be available for CDM- CDM-AR projects being realized there, eligible
AR, in reality, depend upon a more complex set lands identified by the global analysis were
of parameters set within a national, local and site- characterized by existing land use class,
specific socio-economic and ecological context. population density, elevation zone, aridity index,
These conditions go beyond the CDM-AR rules, and productivity classes (figure 4). These factors,
or the biophysical fact that trees grow well on any beyond biophysical suitability, contribute to the
particular piece of land, to include such factors likelihood of land being converted to CDM-AR
15
FIGURE 4.
Socio-ecological characteristics of CDM-AR suitable areas: (a) Existing landuse; (b) Population density (persons/sq
km); (c) Elevation (meters asl); d) Aridity Index (AI); (e) Net Primary Productivity (NPP) (tC/ha/yr); (f) Percentage decrease
in runoff (%) with land use change to CDM-AR; and (g) Decrease in runoff (mm) with land use change to CDM-AR.
16
because they reflect current land use activities, much smaller areas of shrubland and savannah
the number of people who may be dependent on (table 3), reflecting the high population densities
that piece of land, its productivity, and potential and pervasive agricultural production found in
opportunity costs of CDM-AR projects. these regions. Much of the hilly land in South
Asia and the Himalayan foothill areas have
Land Use. Across the five regions, more than canopy cover percentages above the threshold for
50 percent of all the eligible area is classified as forest, although many of these areas are under
within an agricultural land use type, constituting various forms of intensive agricultural production.
more than 364 Mha (figure 4a). This is not More than 52 percent (172 Mha) of the land
surprising, and in line with generally accepted in South America identified as suitable is
assumptions about availability of CDM-AR classified as cropland. An additional 29 Mha is
suitable land. Since the criteria specify that mixed shrubland/grassland, and is likely to be
forested areas are not eligible, and since much under some form of livestock production activity.
deforestation has occurred to make room for Since the Aridity Index was set at a threshold
agriculture, by elimination, agricultural land is left that generally indicates a lack of water stress,
as likely to be available. While intensive these included savannah areas that can be
production sites have been excluded from this considered as more mesic and fairly productive.
analysis, it is likely that other agricultural areas Sub-Saharan Africa has a large amount of
are ideal for optimal tree growth, with deeper savannah (132 Mha) classified as suitable (68
soils, better climate, adequate moisture, and percent), where it is likely that substantial
also meet the CDM-AR criteria, i.e., are not pastoralist and other subsistence livelihood
currently forested. However, the probability for activities are present, even in less populated
much of this area, either currently under areas. Much of this savanna land, although
commercial production, or in subsistence identified as biophysically suitable for tree growth,
farming, to actually convert to CDM-AR is has a very low probability of being converted to
dependent on socio-economic and local food CDM-AR. These semi-arid lands do have a
security issues. In this regard, this estimate potential for agroforestry, and may also have
should be considered a theoretical potential for other options beside tree plantations for
land suitability (Cannell 2003). increasing on-site carbon. Restoration of dry
Both South Asia and Southeast Asia have a forest, for example, addressing losses of these
very high percentage of the land identified as types in the highlands of Ethiopia or Madagascar,
suitable for CDM-AR classified as under although exhibiting very slow growth, can have
agricultural land use types (76 percent), with significant potential for sequestering carbon over
TABLE 3.
CDM-AR suitable land by existing landuse type, by total area (Mha), and percent (%) of the total suitable land,
regionally and globally.
Existing Landuse Type
Mixed Barren/
Shrubland/ Sparsely
Cropland Grassland Savanna Vegetated Total
Region Mha % Mha % Mha % Mha %
East Asia 59 63 20 21 14 15 0 0.1 93
Sub-Sahara Africa 54 28 8 4 132 68 1 0.4 195
South America 172 52 29 9 132 40 1 0.2 333
South Asia 48 76 3 5 12 18 0 0.1 63
Southeast Asia 31 76 3 8 6 16 0 0.2 41
Global 364 50 63 9 296 41 2 0.2 725
17
the long term (IPCC 2000). It is likely, however, of all areas identified in Sub-Sahara Africa have
that slow growing dry forest CDM-AR projects will levels less than 100 persons per sq km. In
require a relatively high price for sequestered contrast, East Asia has 55 percent of its
carbon, and alternative strategies, for example identified areas with population levels above 100
ecotourism, or subsidies, due to their low people/sq km, with 11 percent above 500 people/
financial returns, in order to be viable. sq km. Likewise, South Asia has more than 65
Protected areas and national parks were percent of identified areas with population levels
excluded from this analysis. However, it is above 100 persons/sq km, and 24 percent above
recognized that some degraded areas now 500 persons/sq km. Southeast Asia has 65
designated as protected offer optimal percent of identified areas with population levels
opportunities for reforestation and CDM-AR. A above 100 persons/sq km, and 33 percent with
relevant example is the Mt. Elgon Reforestation less than 25 persons/sq km. Much of the low
Project (FACE 1998), on the slope of Mt. Elgon in population density classes in South America and
eastern Uganda. This National Park was Sub-Sahara Africa are comprised of savanna,
deforested by massive encroachment during the although particularly in South America, substantial
regime of Idi Amin.  Subsequently, the areas of very low population density are classified
government of Uganda reclaimed this area as a as agricultural land use types. In Southeast Asia,
national park, and worked with the FACE (Forests degraded forest areas account for much of the
Absorbing Carbon Emissions) Foundation of the low density areas. In South Asia, cropland
Netherlands to fund reforestation, based on the accounts for the majority of identified areas
carbon sequestration component of the improved across all population density levels. Globally,
ecosystem services provided by the reforestation large areas identified within the savanna land use
and ecosystem restoration. The legal commitment class extend up to density classes of about 200
to permanency provided by the Uganda Wildlife persons/sq km, as influenced by the large
Authority to the National Park provided an ideal amounts of these areas found in South America
opportunity for carbon sequestration. and Sub-Sahara Africa. It seems that except in
Asia, displacement of populations, which is often
Population. Patterns of rural population densities raised as a potential problem for CDM-AR, is not
on suitable land vary widely between regions a major concern.
(figure 4b). Population density is considered here
as a measure of utilization and it is assumed that Elevation. Globally, almost 60 percent of
at high densities less land is likely to be available lands are found below 500 m of
converted to tree plantations.  In addition, it is elevation (table 5), with almost 80 percent below
assumed that in areas of high rural population 1000 m. This trend is generally true for all
densities, competition for food production and regions, except Sub-Sahara Africa (figure 4c),
food security issues will inhibit adoption of CDM- which has about 40 percent below 500 m, and
AR projects.  Globally, more than 50 percent of almost 50 percent between 500 and 1500 m. In
all identified areas have population densities less general, the notion that one would find most of
than 25 people/square kilometer (sq km), that is, these projects in mountainous or sloped areas
have relatively low densities, with more than 35 seems to be discounted at the scale of this
percent with densities less than 5 people/sq km analysis, as demonstrated by relatively little
(table 4). Areas in South America have the lowest available land above 1500 m, less than 10
population levels, with 95 percent of all identified percent globally, with only 20 percent available
areas having less than 100 people/sq km, and above 1000 m. However, it is very likely that on
almost 70 percent less than 5 persons/sq km. hilly, sloped, or mountainous lands, at more local
Sub-Sahara Africa has less empty lands, but still scales, CDM-AR projects may have comparative
has relatively low population densities associated advantages, especially if other ecosystem
with these identified areas. More than 85 percent services are taken into account.
18
TABLE 4.
CDM-AR suitable land by population density class given by area (Mha), and as percent (%) of the total CDM-AR
suitable land, regionally and globally.
Population Density (persons/sq km)
0-10 25 50 100 200 300 500 >500 Total
Region                                   CDM-AR Suitable Land Area (Mha)
East Asia 4 4 7 14 23 13 14 14 93
Sub-Sahara Africa 63 40 30 26 21 5 5 4 195
South America 260 31 15 10 5 4 4 5 333
South Asia 1 0 2 7 18 10 9 16 63
Southeast Asia 5 4 5 7 9 3 3 5 41
                                               CDM-AR Suitable Land Area (Mha)
Global 332 80 59 65 76 35 35 43 725
Region                                   Percent of Total Regional CDM-AR Suitable Land Area (%)
East Asia 4 4 8 15 25 14 15 15
Sub-Sahara Africa 32 21 15 13 11 3 3 2
South America 78 9 5 3 2 1 1 1
South Asia 2 1 3 12 29 15 15 25
Southeast Asia 11 10 13 18 21 8 7 11
                                               Percent of Total Regional CDM-AR Suitable Land Area (%)
Global 46 11 8 9 11 5 5 6
TABLE 5.
CDM-AR suitable land by elevation class, given by area (Mha), and as percent (%) of the total suitable land, regionally
and globally.
Elevation Class (m)
0-500 500-1000 1000-1500 1500-2000 > 2000 Total
Region (Mha) (%) (Mha) (%) (Mha) (%) (Mha) (%) (Mha) (%)
South America 234 70 85 25 9 3 2 0 4 1 333
Sub-Sahara Africa 74 38 50 26 40 20 18 9 12 6 195
South Asia 49 77 11 18 1 2 1 2 1 2 63
Southeast Asia 35 87 3 8 1 3 0 1 0 1 41
East Asia 59 63 14 15 7 8 6 6 8 8 93
Global 451 62 163 23 58 8 27 4 26 2 725
19
Aridity Index. Approximately 30 percent of the Net Primary Productivity. Results obtained from
initially identified areas had values below the a spatial analysis of the NASA MODIS MOD-
optimal threshold value of 0.65 for the Aridity 17A3 NPP product show that lands suitable for
Index, globally (figure 4d). Sites with values CDM-AR generally fall into moderately low to
below 0.65 were considered as sub-optimal for moderate productivity categories (figure 4e),
tree growth, and/or in some cases may not be indicating that higher productivity lands, mainly
suitable for more than mixed shrub and small intensive and irrigated cropping and forested
woody vegetation types. In Africa, 38 percent of areas, were eliminated by the analysis, thus
initially identified areas were below the optimal leaving proportionally large amounts of less
Aridity Index (AI) value of 0.65, and large areas productive land and borderline marginal areas for
in Sub-Saharan Africa, South America (figure 5) afforestation/reforestation. Likewise, many of the
and South Asia were identified within semi-arid most marginal areas were also eliminated by the
zones. While natural forests can be found within Aridity Index criteria, thus giving a generally
these zones, these areas are considered as Gaussian distribution of productivity classes,
marginally suitable for CDM. They may, however, centered on a moderately productive mean.
be utilized for specialized or focused projects, Globally, 88 percent of all available land had a
such as restoration of dry forests. We have NPP below 10 tonnes of carbon/per hectare/per
excluded these areas in our final assessment of year (tC/ha/yr) (table 6). About 75 percent of
total suitable land. available land in Africa and Southeast Asia, and
FIGURE 5.
Aridity Index (AI) was calculated for the entire globe, with aridity maps for South America and Africa shown below.
A threshold value of AI > 0.65 was used as a parameter in the land suitability analysis to delineate CDM-AR suitable
areas.
20
TABLE 6.
CDM-AR suitable land by NPP class given by area (Mha), and as percent (%) of the total suitable land, regionally and
globally.
NPP (tC/ha/yr)
0-2.5 2.5-5.0 5.0-7.5 7.5-10.0 10.0-12.5 12.5-15.0 > 15.0 Total
Region CDM-AR Suitable Land Area (Mha)
East Asia 6.1 62.2 19.3 4.3 1.0 0.4 0.0 93
Sub-Sahara Africa 1.5 9.2 58.9 78.9 36.7 4.0 5.3 195
South America 2.7 45.5 193.9 63.9 14.7 7.2 5.3 333
South Asia 3.9 29.7 23.3 4.1 1.3 0.6 0.3 63
Southeast Asia 0.2 2.7 18.1 9.5 5.6 3.6 1.2 41
CDM-AR Suitable Land Area (Mha)
Global 14 149 314 161 59 16 12 725
Region Percent of Total Regional CDM-AR Suitable Land Area (%)
East Asia 7 67 21 5 1 0 0
Sub-Sahara Africa 1 5 30 41 19 2 3
South America 1 14 58 19 4 2 2
South Asia 6 47 37 7 2 1 1
Southeast Asia 0 7 44 23 14 9 3
Percent of Global CDM-AR Suitable Land Area (%)
Global 2 21 43 22 8 2 2
almost all available land in South America (92 signatory countries. Results of these analyses
percent), South Asia (96 percent) and East Asia are interactively available on-line for each country
(98 percent), indicated a NPP less than 10 tC/ha/ using the ENCOFOR CDM-AR Online Analysis
yr. These results indicate productivity levels Tool, available at http://csi.cgiar.org/encofor/.
consistent with global values (Esser et al. 2000; Results are given on a country by country basis,
Scurlock and Olson 2002) and reflect the with maps, tables, and graphs of the delineated
abundant inclusion of marginal and subsistence area and its socio-ecological characteristics
cropping areas, and lower productivity grassland. presented. In addition, the search tool allows the
user to specify the crown cover density threshold
National Level Land Suitability Analysis and to be used as ‘forest definition’ (Verchot et al.
Socio-Ecological Characteristics 2006), and whether or not to include protected
areas (which includes national parks and other
The land suitability analysis was delineated, bioreserves) within the area deemed suitable for
mapped and tabulated for all Non-Annex I KP afforestation and reforestation.
21
Land required to meet the CDM-AR cap literature survey of tropical tree plantation growth
Including CDM-AR activities into the KP has been rates and the IPCC guidelines (IPCC 2000). The
one of the ‘crunch issues’ in the climate calculation indicates that from 4 to 8 Mha of land
negotiations, and has spawned much debate planted with fast growing tree species will easily
(Noble and Scholes 2001). In addition to the satisfy the total allowable demand for CERs.
basic controversy with regards to the Assumptions incorporated into this estimate
effectiveness of CDM-AR to mitigate GHG include accounting for baseline and the lower
emissions, controversial issues include productivity of marginal or degraded areas. It is
measurement of carbon sequestration, further assumed that many of these projects,
permanence, leakage, land conflicts and which are likely to have goals beyond maximizing
environmental considerations (Schlamadinger and profitability, are likely to be less productive than
Marland 2000; Torvanger et al. 2001), as well as typical intensively managed commercial tree
various technical and scientific aspects of carbon plantations as they are found in the tropics.
sequestration in agriculture and forestry examined This is a relatively small figure, representing
by the Special IPCC Report (IPCC 2000) less that 1-2 percent of the area we have
commissioned by the Subsidiary Body for identified as suitable. CDM-AR is likely to be
Scientific and Technological Advice (SBSTA of relatively small compared to globally suitable area
the UNFCCC), after the Sixth Conference of the estimates, and be geographically dispersed, both
Parties (COP-6) held in Bonn in 2001. nationally and globally.  Although small compared
Afforestation and reforestation are currently the to the total global suitable area estimate, the total
only eligible LULUCF activities under Article 12 amount of land, and the potential funds made
(UNFCCC 2002a; UNFCCC 2002b), specifically available for development, can be significant,
excluding activities such as avoidance of both locally and nationally, depending upon rate of
deforestation, improved forest management, or adoption, and especially dependent upon the
agricultural activities that build up carbon, such market price for CERs.
as conservation farming. Eligible projects have to
represent a real land use change from non-forest Water use impact of CDM-AR
into forest, or agroforestry, thus preventing current
forests being converted into plantations (Smith Land use changes resulting from the adoption of
and Scherr 2002). CDM-AR involve alterations of the hydrological
In response to widespread concerns that cycle, both on flows of water and sediment and in
CDM sink projects would impact negatively on situ vapor flow. Both, the relative impact on water
CO cycles and absolute change in the quantity of
2 emission reduction aims (Greenpeace 2003),
a cap on CDM-AR emission reduction offsets was water moving away from the site either as vapor
set at one percent of the total global emission or runoff, were quantified and mapped in this
reduction target. The limit on the use of sink analysis.  Together they indicate that large areas
projects under the CDM implies that the annual deemed suitable for CDM-AR would exhibit
flow of Certified Emissions Reductions (CERs) significant increases in vapor flow (figure 6) and/
from afforestation and reforestation under Article or substantial decreases in runoff (figure 7). This
12 has an upper limit of 32.6 megatonnes of is particularly evident in drier areas, the semi-arid
Carbon (Mt C), representing 119.6 megatonnes of tropics, and in conversion from grasslands and
Carbon Dioxide (Mt CO2) equivalents, based on subsistence agriculture. Significant variation
UNFCCC emission figures (Kolshus 2001). In amongst biomes and bioclimatic zones is evident.
order to make a rough estimate of the amount of However, almost 20 percent (144 Mha) of all
land that would be required to fully meet this cap, suitable land showed little or no impact on runoff
we used an averaged estimate for annual carbon with another 28 percent (210 Mha) showing only
sequestration (4 to 8 tC/ha/yr), based on a moderate impact (table 7).
22
FIGURE 6.
Increases in vapor flow resulting from landuse change to CDM-AR, are given both in absolute terms (mm), and as the
percentage increase (%) from existing landuse. Vapor flow includes both the AET and Int components of the water
balance model.
23
FIGURE 7.
Decreases in runoff resulting from landuse change to CDM-AR, are given both in absolute terms (mm), and as the
percentage decrease (%) from existing landuse.
24
TABLE 7.
Decrease in total runoff (mm) and percent decrease (%) in total runoff with landuse change to CDM-AR on suitable
land, regionally and globally.
Decrease in Runoff (mm)
0-50 50-100 100-150 150-200 200-250 250-300 300-400 > 400
Region (Mha)
East Asia 16 14 10 16 18 7 2 1
Sub-Sahara Africa 15 19 45 67 30 12 9 2
South America 38 42 57 81 72 38 27 5
South Asia 0 1 2 8 9 12 27 6
Southeast Asia 0 1 2 3 3 3 8 11
(Mha)
Global 69 76 116 175 132 73 72 25
Decrease in Runoff as a Percent of Total (%)
0-20 20-40 40-60 60-80 80-100
Region (Mha)
East Asia 6 10 25 21 22
Sub-Sahara Africa 0 13 11 3 2
South America 3 12 22 19 9
South Asia 7 33 58 53 48
Southeast Asia 11 48 94 87 119
(Mha)
Global 28 116 210 183 200
Taken together 50 percent of all suitable land scale or even major regional impacts would be
showed a decrease in runoff of less than 60 evident in the aggregated statistics. Further, with
percent (figure 4f). About 27 percent (200 Mha) is the cap on CDM-AR at one percent, estimated by
in the highest impact class exhibiting an 80-100 this study to be satisfied by at most conversion
percent decrease in runoff. Altogether, almost 60 of a mere 2 percent of available land, direct
percent showed a decrease of less than 200 mm, impacts of CDM-AR at the global and regional
with only slightly more than 13 percent showing a scales are unlikely. However, significant changes
decrease of more than 300 mm (figure 4g). Since in CDM rules affecting the number of carbon sink
it is reasonable to assume that only a small projects, or amount of land which will eventually
proportion of these lands would be converted to be under CDM-AR, should take into account
forestry land use types, it is unlikely that global these potential impacts on the hydrological cycle.
25
Local Scale Water Use Impact of CDM-AR during the rainy season is a major problem for the
adjacent city of Cochabamba, thus decreased
At the local and project level, impacts were runoff and lowered water tables as demonstrated
estimated to be substantial and important. Land in this study are considered positive. Thus, tree
use change to CDM-AR in all of the four study planting for the Tunari site is shown to be an
sites showed strong spatial variations of runoff effective means to provide multiple benefits such
and changes in SWC. While reduced runoff is as conservation and flood mitigation. In the
clearly linked to reduced downstream water Guamote case study, in the highland Sierras of
supply, variation in SWC is also important Ecuador (figure 10), the water implications of
because it implies a likely associated variation in afforestation with pine trees is already a
groundwater tables. It is usually assumed that controversial issue (Farley et al. 2004). In
most, if not all, base flow is supplied by addition to a large decrease in runoff (54 percent),
groundwater circulation, and initially by downward there also appears to be a significant impact on
flows associated with SWC above field capacity. the soil water content (decrease of 32 percent),
Streams that receive large proportions of their indicating a likelihood of decreasing water table
flow as groundwater base flow tend to have levels over time. Increases in AET and total
relatively low temporal flow variability and hence vapor flows are relatively small, since this
provide a more reliable source of water for system is already water limited under current land
various water-resource purposes (Dingman 1993). use. As predicted in this case, common
All four sites showed a marked reduction in consequences of afforestation projects using fast-
runoff, with both on-site and off-site implications growing conifers are decreased levels of stream
(table 8). On the humid lowland tropical Amazon flow, both over the entire year (Swank and
site in Chapare, Bolivia (figure 8), the impact of Douglass 1974) and during the dry season
the reduction was minimal, since precipitation is (Vincent 1995). Likewise, the reduction in runoff
high and not a limiting factor. By contrast, the associated with conversion of pasture to mixed
drier high elevation Tunari site in Bolivia (figure 9) tropical indigenous agroforestry in coastal
showed significant decrease (28 percent) in Ecuador (figure 11) was relatively large (47
runoff. There was relatively little impact on soil percent). However, again in this case, the
water content since these denuded slopes already generally higher level of precipitation and the
have a very low water holding capacity under the site’s downstream location within the catchment,
existing land use. At this site, recurrent flooding minimized the importance of the decrease in
due to excessive runoff from eroded slopes runoff.
TABLE 8.
Results of water balance model applied at local scale for four case studies. Project area represent the total area
allocated to the project, and CDM-AR area is the total area within the project area suitable for CDM-AR. Vapor flow
is given as the sum of AET and Int, in order to represent total ET, and is presented as the percent increase resulting
from landuse change to CDM-AR. Runoff and SWC are given as the percent decrease resulting from landuse
change to CDM-AR.
CDM-AR Project Project CDM-AR Precip Aridity Vapor Flow Runoff SWC
Area Area Index Increase Decrease Decrease
(ha) (ha) (mm/yr) (Mean AI) (%) (%) (%)
Tunari NP, Bolivia 32,142 9,873 900 0.8 7.1 27.7 7.3
Chapare, Bolivia 40,604 11,077 3,000 1.8 15.1 12.4 1.1
Guamote, Ecuador 15,104 13,327 700 0.6 4.7 54.0 32.0
Coastal Ecuador 41,878 26,564 1,300 0.9 23.4 47.4 13.4
Note: Precip=Precipitation.
26
FIGURE  8.
Chapare Case Study: (a) CDM-AR suitable land; (b) increase in vapor flow (AET and Int) with landuse change to
CDM-AR; (c) decrease in SWC with landuse change to CDM-AR; (d) decrease in Runoff with landuse change to
CDM-AR; and (e) representative view of the project area, showing a mixed farming landscape typical of this area in
the Bolivian Amazon.
27
FIGURE  9.
Tunari Case Study: (a) CDM-AR suitable land; (b) increase in vapor flow (AET and Int) with landuse change to CDM-
AR; (c) decrease in SWC with landuse change to CDM-AR; (d) decrease in Runoff with landuse change to CDM-AR;
and (e) view of reforestation with pine in the Tunari National Park, with the city of Cochabamba below.
28
FIGURE  10.
Guamote Case Study: (a) CDM-AR suitable land; (b) increase in vapor flow (AET and Int) with landuse change to
CDM-AR; (c) decrease in SWC with landuse change to CDM-AR; (d) decrease in Runoff with landuse change to
CDM-AR; and (e) community-owned afforestation projects in one of the poorest regions in the highlands of Ecuador.
29
FIGURE 11.
Coastal Ecuador Case Study: (a) CDM-AR suitable land; (b) increase in vapor flow (AET and Int) with landuse
change to CDM-AR; (c) decrease in SWC with landuse change to CDM-AR; (d) decrease in Runoff with landuse
change to CDM-AR; and (e) pastures throughout the humid tropics offer opportunities for increasing carbon baselines.
30
In general, the results indicate that, studies illustrates that both local effects, and
although impacts may not be discernable at the desired outcomes, are highly site specific and
global or regional level, CDM-AR projects have highlights the importance of considering
large and significant local impacts on water hydrologic implications of land use change,
use, with both on-site and downstream when evaluating, planning and implementing
implications. Investigation of these four case CDM-AR.
Conclusion
This report highlights that there is an abundance of constitute an option for significantly increasing
land for, and potentially significant impacts the carbon sequestration within rural and
resulting from, climate change mitigation agricultural landscapes, while contributing
measures, particularly on the hydrologic cycle. The positively to increased food and household
global impact of redistribution of water use driven security. CDM-AR rules do not currently
by agriculture and land use change, of which encourage, or make it easy to promote these
CDM-AR can be a contributing factor, is a major types of small scale, small holder, less intensive
component of ongoing global change, with high approaches, and it is more likely that much of
significance in terms of impact on climate change CDM-AR projects will be in the form of fast-
processes. The CDM-AR hydrological impact growing timber plantations. As such, there are
analysis shows significant impacts on local and indeed both national and local food security
regional hydrologic cycles, although they are not considerations that must be taken into account in
evident at regional or global scale under current proposing CDM-AR development activities. In
rules which limit the amount of sink projects to a many areas, food security may not be an issue,
one percent cap. If the cap on CDM-AR were certainly not regionally or nationally. However, in
raised to compensate for a substantially greater areas with insecure or highly unequal tenure
offset of carbon emission through sink projects, it rights, in systems where large numbers of tenant
is suggested that it will be increasingly important farmers may be displaced due to the lower labor
to consider implications on local to regional water requirements of forestry activities, or access to
resources. Although not currently of this same land by indigenous communities may be lost, the
magnitude (i.e., under the one percent cap), this displacement of subsistence farming activities
important dimension of CDM-AR should be may be of high concern (Smith and Scherr 2002).
formally articulated and taken into account within In contrast, examples of poplar-based
the CDM-AR guidelines, especially when agroforestry from northern India (Gupta et al.
addressing issues of sustainability, local 2005) demonstrate that small-scale wood
communities, and food security. plantations and agroforestry can substantially
The potential for small farmers and increase carbon stocks with significant positive
communities to participate in CDM-AR has been benefits for rural communities. In this case,
highlighted and promoted by developing countries where intensively cultivated irrigated areas are
and NGOs. In particular, the adoption of being converted to agroforestry, afforestation
agroforestry type practices has been put forward provides added security to small farmer
as a way for smaller farmers and communities to livelihoods by offering alternate production
participate in CDM-AR projects. This may opportunities and diversification.
31
The afforestation of upland catchments with locations which minimize negative aspects.
fast growing plantations can have significant Projects can even capitalize on the positive
impact on in situ water use, with consequent aspects of these potential impacts, for instance,
impacts on water availability downstream. in reducing recurrent flooding, or sediment
Generally, CDM-AR results in an increase of AET, transfer.
or ‘green water’ vapor flows, increased on-site It is evident that the supply of potentially
water use, and decreased movement of water and available land, and consequently the potential
sediments off-site. However, whether this is a supply of carbon which can be sequestered, is far
positive or negative impact on water resources, greater than the current cap on CDM-AR credits.
water management, soil and land conservation, It is likely that CDM-AR, and possibly other
biodiversity, and/or downstream food security, is carbon sink approaches, will play a larger,
highly site specific, and dependent upon climate, increasingly more important role in the future,
soil types, topography, land uses, population most probably starting in the second KP
densities, existing infrastructures, and tradeoffs commitment period. This analysis shows that the
with coexisting demands for water. Whereas trees potential for carbon sequestration by sink projects
do use more water than many other vegetation is great. Current negotiations also bring up the
forms and most crops, this analysis has shown prospect of innovative approaches, which could
that the variability in response is highly include avoided deforestation, and restoration of
dependent on the specific ecological degraded forests, so that credits available from
characteristics of the site, and that globally, there sink projects will increase. In addition, we
are large areas of land where impacts of CDM-AR highlighted here the ‘hidden’ water dimension
on water resources and food security will be associated with climate change mitigation efforts
minimal. On a national and local basis, the that can be found in many of the other global
selection of CDM-AR sites can take into treaties and conventions addressing the various
consideration these specific hydrologic and socio- contemporary environmental and global issues.
ecological aspects, to evaluate increased green Articulating these ‘secondary effects’ on the
water vapor flows and associated decreases in hydrologic cycle is essential if we are to address
runoff, and to identify optimal conditions and these global concerns in a holistic fashion.
32
Literature Cited
Alcamo, J.; Doll, P.; Kaspar, F.; and Siebert, S. 1997. Global change and global scenarios of water use and
availability: An application of WaterGAP 1.0. Kassel, Germany: CESR (Center for Environmental Systems
Research), University of Kassel.
Alcamo, J.; Henrichs, T.; and Rosch, T. 2000. World water in 2025: Global modeling and scenario analysis. In:
World Water Scenarios Analyses, ed. F. R. Rijsberman. (World Water Council, Marseille).
Allen, R. G.; Pereira, L. S.; Raes, D.; and Smith, M. 1998. Crop evapotranspiration - guidelines for computing crop
water requirements. FAO Irrigation and Drainage Paper 56. Rome: FAO (Food and Agriculture Organization
of the United Nations).
Aylward, B.; Echeverria, J.; Gonzales, A. F.; Porras, I.; Allen, K.; and Mejias, R. 1998. Economic incentives for
watershed protection: A case study of Lake Arenal, Costa Rica. p.140 in: Final report to the Government of
the Netherlands under the program of Collaborative Research in the Economics of Environment and
Development (CREED). London: IIED (International Institute for Environment and Development), TSC (Tropical
Science Center) and the International Center for Economic Policy, National University at Heredia (CINPE).
Berndes, G. 2002. Bioenergy and water – the implications of large-scale bioenergy production for water use and
supply. Global Environmental Change 12: 253-271.
Bosch, J. M.; and Hewlett, J. D. 1982. A review of catchment experiments to determine the effect of vegetation
changes on water yield and evapotranspiration. Journal of Hydrology 55(1-4): 3-23.
Brown, S.; Sathaye, J.; Cannell, M.; and Kauppi, P. 1996. Management of forests for mitigation of greenhouse
gas emissions. Chap. 24 in Climate change 1995: Impacts, adaptations and mitigation of climate change:
Scientific-Technical analyses. Contribution of working group II to the second assessment report of the
Intergovernmental Panel on Climate Change, eds. R. T. Watson; M. C. Zinyowera; and R. H. Moss. Cambridge
University Press, Cambridge, United Kingdom and New York, NY, USA. pp. 773-797.
Brown, S.; Swingland, I. R.; Hanbury-Tenison, R.; Prance, G. T.; and Myers, N. 2002. Changes in the use and
management of forests for abating carbon emissions: issues and challenges under the Kyoto Protocol. In:
Capturing Carbon and Conserving Biodiversity, ed. I. R. Swingland. London, UK: Earthscan Publications Ltd.
pp. 42-55.
Calder, I. R. 1992. Water use in eucalypts - a review. In: Growth and Water Use of Forest Plantations, eds. I. R.
Calder; R. L. Hall; and P. G. Adlard. New York: J. Wiley. 167-179.
Calder, I. R. 2000. Land use impacts on water resources. Land-Water Linkages in Rural Watersheds Electronic
Workshop, Background Paper No. 1. September 18 – October 27, FAO.
Cannell, M. 2003. Carbon sequestration and biomass energy offset: theoretical, potential and achievable capacities
globally, in Europe and the UK. Biomass and Bioenergy 24(2003): 97-116.
CIESIN (Center for International Earth Science Information Network), Columbia University; and CIAT (Centro
Internacional de Agricultura Tropical). 2005. Gridded Population of the World Version 3 (GPWv3). Palisades,
NY: Socioeconomic Data and Applications Center (SEDAC), Columbia University. Available at http://
sedac.ciesin.columbia.edu/gpw.
CIFOR (Center for International Forestry Research) and FAO (Food and Agriculture Organization of the United
Nations). 2005. Forests and floods: drowning in fiction or thriving on facts? Forest Perspectives. No. 2. 38p.
Costello, L. R.; and Jones, K. S. 2000. Water Use Classification of Landscape Species (WUCOLS III). A guide
to estimating irrigation water needs of landscape plantings in California. Sacramento, California: Department
of Water Resources.
33
Dingman, S. L. 1993. Physical hydrology. Prentice Hall. New Jersey, USA.
Droogers, P.; and Allen, R. G. 2002. Estimating reference evapotranspiration under inaccurate data conditions.
Irrigation and Drainage Systems 16(1): 33-45.
Dyck, S. 1983. Overview on the present status of the concepts of water balance models. IAHS Publication No.
148. Wallingford, UK. 3-19.
Esser, G.; Lieth, H. F. H.; Scurlock, J. M. O.; and Olson, R. J. 2000. Osnabrück net primary productivity data set.
Ecology 81: 1177.
FACE (Forests Absorbing Carbon Emissions). 1998. Annual Report 1997. Arnhem, Netherlands: FACE Foundation.
28 pp.
Falkenmark, M. 1995. Land and water integration and river basin management. In: FAO Land and Water Bulletin
Number 1. Rome, Italy: FAO. pp 15-16.
Falkenmark, M.; and Lannerstad, M. 2004. Consumptive water use to feed humanity – curing a blind spot.
Hydrology and Earth System Sciences Discussion 1: 7-40.
FAO (Food and Agriculture Organization of the United Nations). 1995. Digital Soil Map of the World (DSMW):
based on the FAO/UNESCO Soil Map of the World, original scale 1:5 000 000. FAO-AGL (Land and Water
Development Division). Rome.
FAO. 2001a. FAOCLIM - world-wide agroclimatic data. Version 2. FAO – SDRN (Environment and Natural
Resources Service). Rome.
FAO. 2001b. Crops and drops: Making the best use of water for agriculture. FAO. Rome.
FAO. 2003a. Agriculture, food and water: A contribution to the “World Water Development Report”. FAO. Rome.
FAO. 2003b. Unlocking the water potential of agriculture. FAO. Rome.
Farley, K. A.; Kelly, E. F.; and Hofstede, R. G. M. 2004. Soil organic carbon and water retention after conversion
of grasslands to pine plantations in the Ecuadorian Andes. Ecosystems 7(7): 729-739.
Farrington, P.; and Bartle, G. A. 1991. Recharge beneath a Banksia woodland and a Pinus pinaster plantation
on coastal deep sands in south Western Australia. Forest Ecology and Management 40: 101-118.
Forner, C.; and Jotzo, F. 2002. Future restrictions for sinks in the CDM: How about a cap on supply?  Climate
Policy 2: 353-365.
Gedney, N.; Cox, P. M.; Betts, R. A.; Boucher, O.; Huntingford, C.; and Stott, P. A. 2006. Detection of a carbon
dioxide effect in continental river runoff records. Nature 439: 835-838.
Gordon, L. J.; Steffen, W.; Jonsson, B. F.; Folke, C.; Falkenmark, M.; and Johannessen, A. 2005. Human
modification of global water vapor flows from the land surface. PNAS (Proceedings of the National Academy
of Sciences of the United States of America). 102: 7612-7617.
Greenpeace. 2003. Sinks of the CDM: After the climate, will biodiversity go down the drain? Greenpeace Briefing
Paper, May 30, 2003. Bonn, Germany: Greenpeace International.
Gupta, D. C.; Zomer, R. J.; and Bossio, D. A. 2005. Poplar agroforestry in India: Trends and impacts. IWMI Project
Report. Colombo, Sri Lanka: International Water Management Institute.
Hamilton, E. L.; and Rowe, P. B. 1949. Rainfall interception by chaparral in California. Division of Forestry,
Sacramento, Calif.
Hansen, M.; DeFries, R.; Townshend, J. R.; Carroll, M.; Dimiceli, C.; and Sohlberg, R. 2003. 500m MODIS
Vegetation Continuous Fields. College Park, Maryland: The Global Land Cover Facility.
Hargreaves, G. H. 1994. Defining and using reference evapotranspiration. Journal of Irrigation and Drainage
Engineering. ASCE. 120(6): 1132-1139.
34
Hargreaves, G. H.; and Allen, R. G. 2003. History and evaluation of Hargreaves evapotranspiration equation.
Journal of Irrigation and Drainage Engineering. ASCE. 129(1): 53-63.
Hargreaves, G. L.; Hargreaves, G. H.; and Riley, J. P. 1985. Irrigation water requirements for Senegal River Basin.
Journal of Irrigation and Drainage Engineering. ASCE. 111(3): 265-275.
Heinsch, F. A.; Zhao, M.; Running, S. W.; and Kimball, J. S. 2005. Evaluation of remote sensing based terrestrial
productivity from MODIS using regional tower eddy flux network observations. IEEE Transactions on
Geosciences and Remote Sensing. 44: 1908-1925.
Heuvelmans, G.; Garcia, J.; Muys, B.; Feyen, J.; and Coppin, P. 2005. Modelling the water balance with SWAT
as part of the land use impact evaluation in a life cycle study of carbon sequestration scenarios. Hydrological
Processes 19(3): 729-748.
Hijmans, R. J.; Cameron, S. E.; Parra, J. L.; Jones, P. G.; and Jarvis, A. 2004. The WorldClim interpolated global
terrestrial climate surfaces. Version 1.3. Downloaded at: http://biogeo.berkeley.edu/worldclim/worldclim.htm.
Holland, H. D. 1978. The Chemistry of the Atmosphere and Oceans. New York, Wiley.
Huq, S. 2002. Applying sustainable development criteria to CDM projects: PCF experience. PCFplus Report 10.
Washington, DC: World Bank.
IPCC (Intergovernmental Panel on Climate Change). 1996. Climate change 1995: Impacts, adaptation, and
mitigation of climate change: Scientific-Technical analysis. The contribution of working group II to the second
assessment report of the Intergovernmental Panel on Climate Change. eds. R. T. Watson; M. C. Zinyowera;
and R. H. Moss. Cambridge, UK: Cambridge University Press.
IPCC. 2000. IPCC Special Report on Land Use, Land-Use Change and Forestry. A special report of the
Intergovernmental Panel on Climate Change. Approved at IPCC Plenary XVI (Montreal, May 1-8, 2000). IPCC
Secretariat, c/o World Meteorological Organization, Geneva, Switzerland. At http://www.ipcc.ch/.
IPCC. 2001. Climate Change 2001, Impacts, Adaptation, and Vulnerability. Contribution of Working Group II to
the IPCC Third Assessment Report (TAR). Cambridge, UK: Cambridge University Press. (in press).
IWMI (International Water Management Institute). 2000. World water supply and demand in 2025. In: World Water
Scenario Analyses. ed. F. R. Rijsberman. World Water Council, Marseille.
Jackson, R. B.; Jobbágy, E. G.; Avissar, R.; Roy, S. B.; Barrett, D. J.; Cook, C. W.; Farley, K. A.; le Maitre, D. C.;
McCarl, B. A.; and Murray, B. C. 2005. Trading water for carbon with biological carbon sequestration. Science.
310(5756): 1944-1947.
Jensen, M. E.; Burman, R. D.; and Allen, R. G. 1990. Evapotranspiration and irrigation water requirements. ASCE
Manuals and Reports on Engineering Practice. New York: American Society of Civil Engineers.
Jung, M. 2003. The role of forestry sinks in the CDM - Analysing the effects of policy decisions on the carbon
market. HWWA Discussion Paper 241. Hamburgisches Welt-Wirtschafts-Archiv (HWWA). Hamburg, Germany:
Hamburg Institute of International Economics.
Jung, M. 2005. Host country attractiveness for CDM non-sink projects. HWWA Discussion Paper 312.
Hamburgisches Welt-Wirtschafts-Archiv (HWWA). Hamburg, Germany: Hamburg Institute of International
Economics.
Kolshus, H. H. 2001. Carbon sequestration in sinks: An overview of potential and costs. CICERO Working Paper
2001. Blindern, Norway: Center for International Climate and Environmental Research - Oslo.
Kolshus, H. H.; Vevatne, J.; Torvanger, A.; and Aunan, K. 2001. Can the Clean Development Mechanism attain
both cost-effectiveness and sustainable development objectives? CICERO Working Paper 2001:8. Blindern,
Norway: Center for International Climate and Environmental Research - Oslo.
35
Korner, C.; and Paulsen, J. 2004. A world-wide study of high altitude treeline temperatures. Journal of
Biogeography 31(5): 713-732.
L’vovich, M. I.; and White, G. F. 1990. Use and transformations of terrestrial water systems, In: The earth as
transformed by human action: Global and regional changes in the biosphere over the past 300 years. ed. B.
L. Turner II. Cambridge, UK: Cambridge University press.
Le Maitre, D. C.; Scott, D. F.; and Colvin, C. 1999. A review of information on interactions between vegetation
and groundwater. Water SA 25(2): 137-152.
Maidment, D. R. 1992. Handbook of hydrology. New York, McGraw-Hill.
Matthews, D. 2006. The water cycle freshens up. Nature 439: 793-794.
Nilsson, S.; and Schopfhauser, W. 1995. The carbon-sequestration potential of a global afforestation program.
Climatic Change 30(3): 267-293.
NIMA (National Imagery and Mapping Agency). 1997. VectorMap (VMAP) Level 1. Source: National GeoSpatial
Agency, U. S. Dept. of Defense.
Noble, I. R.; and Scholes, R. J. 2001. Sinks and the Kyoto Protocol. Climate Policy 1:1-23.
Nordhaus, W. D. 1991. The cost of slowing climate change: A survey. The Energy Journal 12(1): 37-65.
Raskin, P.; Gleick, P.; Kirshen, P.; Pontius, G.; and Strzepek, K. 1997. Water futures: Assessment of long-range
patterns and prospects. Stockholm: Stockholm Environment Institute.
Rijsberman, F.; Bossio, D.; and Noble, A. D. 2006. The impact of land degradation and improved land
management on water productivity. In: International conference of soil, water and environmental quality – issues
and strategies, New Delhi, India, January 28 - February 1, 2005.
Rockstrom, J.; Gordon, L.; Folke, C.; Falkenmark, M.; and Engwall, M. 1999. Linkages among water vapor flows,
food production, and terrestrial ecosystem services. Conservation Ecology 3(2): 5.
Running, S. W.; Thornton, P. E.; Nemani, R. R.; and Glassy, J. M. 2000. Global terrestrial gross and net primary
productivity from the earth observing system. In: Methods in Ecosystem Science. eds. O. Sala; R. Jackson; and
H. Mooney. New York: Springer-Verlag.
Sathaye, J.; and Ravindranath, N. H. 1998. Climate change mitigation in the energy and forestry sectors of
developing countries. Annual Review of Energy Environment. 23: 387-437.
Schenk, H. J.; and Jackson, R. B. 2002. The global biogeography of roots. Ecological Monographs 72:311-328.
Source: F. G. Hall; G. Collatz; S. Los; E. Brown de Colstoun; and D. Landis. eds. ISLSCP Initiative II. NASA.
DVD/CD-ROM. NASA, 2005.
Schlamadinger, B.; and Marland, G. 2000. Land use and global climate change: Forests, land management and
the Kyoto Protocol. Arlington, Vermont: Pew Centre on Climate Change.
Schroth, G.; da Silva, L. F.; Wolf, M. A.; Teixeira, W. G.; and Zech, W. 1999. Distribution of throughfall and stemflow
in multi-strata agroforestry, perennial monoculture, fallow and primary forest in central Amazonia, Brazil.
Hydrological Processes 13: 1423-1436.
Schulze, E. D. 1982. Plant life forms and their carbon, water and nutrient relations. In: Physiological Plant Ecology
II. Water Relations and Carbon Assimilation. eds. O. L. Lange; C. B. Osmond; and V. Ziegler. Berlin, Germany:
Springer-Verlag. 615-676.
Scurlock, J. M.; and Olson, R. J. 2002. Terrestrial net primary productivity - A brief history and a new worldwide
database. Environmental Reviews. 10(2): 91-109.
Seckler, D.; Amarasinghe, U.; Molden, D. J.; de Silva, R.; and Barker, R. 1998. World water demand and supply,
1990 to 2025: Scenarios and issues. IWMI Research Report 19. Colombo, Sri Lanka: International Water
Management Institute.
36
Sedjo, R.; and Solomon, A. 1989. Climate and forests. Paper for Workshop on Controlling and Adapting to
Greenhouse Forcing, held by Environmental Protection Agency and National Academy of Sciences 14-15 June,
1988, Washington, DC. In: Greenhouse warming: abatement and adaptation. eds. N. Rosenberg; W. E.
Easterling; P. R. Crosson; and J. Darmstadter. Washington, DC: Resources for the Future.
Shiklomanov, I. A. 1991. The world’s water resources. In: Proc. International Symposium to Commemorate 25
Years of the IHP. Paris: UNESCO/IHP. pp. 93-126.
Shiklomanov, I. A. 1998. World water resources: An appraisal for the 21st century. IHP Report. Paris: UNESCO.
Smakhtin, V.; Revenga, C.; and Doll, P. 2004. Taking into account environmental water requirements in global-
scale water resources assessments. Comprehensive Asessment of Water Management in Agriculture Research
Report 2. Colombo, Sri Lanka: International Water Management Institute.
Smith, J.; and Scherr, S. 2002. Forests, carbon and local livelihoods: An assessment of opportunities and policy
recommendations. CIFOR, Bogor, Indonesia.
Swank, W. T.; and Douglass, N. H. 1974. Streamflow greatly reduced by converting deciduous hardwood stands
to pine. Science 185: 857-859.
Tate, K. W. 1996. Interception on rangeland watersheds. Fact Sheet No. 36. Univ. of Calif. Cooperative Extension
and USDA Natural Resources Conservation Service, Rangeland Watershed Program.
Thornthwaite, C. 1948. An approach toward a rational classification of climate. Geographical Review. 38(1): 55-94.
Thornthwaite, C.; and Mather, J. 1955. The water balance. Publ. Climatol. 8(1): 1-104.
Thurow, T. L.; Blackburn, W. H.; and Taylor, C. A. 1987. Rainfall interception by midgrass, shortgrass and live oak
mottes. Journal of Range Management 40: 455-460.
Torvanger, A.; Alfsen, K. H.; Kolshus, H. H.; and Sygna, L. 2001. The state of climate policy research and climate
policy. CICERO Working Paper 2001:2. Blindern, Norway: Center for International Climate and Environmental
Research - Oslo.
Trexler, M.; and Haugen, C. 1995. Keeping it green: Tropical forestry opportunities for mitigating climate change.
World Resources Institute and Environmental Protection Agency. Washington, DC.
Turner, D. P.; Ritts, W. D.; Cohen, W. B.; Gower, S. T.; Zhao, M.; Running, S. W.; Wofsy, S. C.; Urbanski, S.; Dunn,
A. L.; and Munger, J. W. 2003. Scaling Gross Primary Production (GPP) over boreal and deciduous forest
landscapes in support of MODIS GPP product validation. Remote Sensing of the Environment. 88: 256-271.
U. S. Bureau of Reclamation. 2005. Agrimet Data System.
UNEP (United Nations Environment Programme). 1997. World atlas of desertification 2ED. London: UNEP.
UNFCCC (United Nations Framework Convention on Climate Change). 2002a. Report of the Conference of the
Parties on its seventh session, held in Marrakech from October 29 - November 10, 2001. Addendum Part Two:
Action Taken by the Conference of the Parties Volume I (FCCC/CP/2001/13/Add. 1), United Nations Framework
Convention on Climate Change Secretariat, Bonn, Germany.
UNFCCC. 2002b. Report of the Conference of the Parties on its seventh session, held in Marrakech from October
29 - November 10, 2001. Addendum Part Two: Action Taken by the Conference of the Parties Volume II (FCCC/
CP/2001/13/Add. 1), United Nations Framework Convention on Climate Change Secretariat, Bonn, Germany.
USGS (United States Geological Survey). 1993. The Global Ecosystem Land Cover Characterization Database
version 2.0, was retrieved from http://edcdaac.usgs.gov/glcc/glcc.html. United States Geological Survey Earth
Resources Observation System Distributed Active Archive Center. Viewed 2005.
USGS. 2004. Reprocessing by the GLCF. (1, 3, 30) Arc Second SRTM Elevation, Reprocessed to GeoTIFF. College
Park, Maryland: The Global Land Cover Facility. Version 1.0.
37
Verchot, L. V.; Zomer, R. J.; van Straaten, O.; and Muys, B. 2006. Implications of country-level decisions on the
specification of crown cover in the definition of forests for land area eligible for afforestation and reforestation
activities in the CDM. Journal of Climate Change. In press.
Vincent, J. R. 1995. Timber trade, economics, and tropical forest management. Ecology, Conservation and
Management of Southeast Asia Rainforests. New Haven, CN, USA: Yale University Press. 241-261.
Vorosmarty, C. J.; Green, P.; Salisbury, J.; and Lammers, R. B. 2000. Global water resources: vulnerability from
climate change and population growth. Science 289: 284-288.
Vrolijk, C.; and Grubb, M. 2001. Quantifying Kyoto: How will COP6 decisions affect the market? Workshop Report,
Royal Institute of International Affairs. August 2000, Chatham House, London.
Wallace, J. S. 2000. Increasing agricultural water efficiency to meet future food production. Agriculture, Ecosystems
and Environment. 82: 105-119.
Wallace, J. S.; and Gregory, P. J. 2002. Water resources and their use in food production. Aquatic Sciences
64(2002): 363-375.
WDPA (World Database on Protected Areas) Consortium. 2004. Copyright World Conservation Union (IUCN) and
United Nations Environment Programme - World Conservation Monitoring Centre (UNEP-WCMC), 2004. Source
for this dataset was the Global Land Cover Facility: http://glcf.umiacs.umd.edu/.
WHO (World Health Organization). 2006. Water Sanitation and Health Website: http://www.who.int/
water_sanitation_health/hygiene/en/.
Wilm, H. G. 1957. The influence of forest vegetation on water and soil. Unasylva 11(4).
Winjum, J. K.; Brown, S.; and Schlamadinger, B. 1998. Forest harvests and wood products: sources and sinks
of atmospheric carbon dioxide. Forest Science 44: 272-284.
Yamagata, Y.; and Alexandrov, G. A. 2001. Would forestation alleviate the burden of emission reduction? An
assessment of the future carbon sink from ARD activities. Climate Policy 1: 27-40.
Young, J. A.; Evans, R. A.; and Easi, D. A. 1984. Stemflow on western juniper (Juniprus occidentalis) trees. Weed
Science 32: 320-327.
Zhao, M.; Heinsch, F. A.; Nemani, R. R.; and Running, S. W. 2005. Improvements of the MODIS terrestrial gross
and net primary production global data set. Remote Sensing of Environment 95(2): 164-176.
Zomer, R. J.; van Straaten, O.; and Verchot, L. V. 2004. Manual for: bio-physical characterization and land suitability
analysis for CDM-AR projects. ENCOFOR Working Paper. ENCOFOR Project: http://www.joanneum.at/encofor.
Zomer, R. J.; van Straaten, O.; and Verchot, L. V. 2006. Land suitability modeling and project scenario analysis
for CDM-AR. ENCOFOR Working Paper. ENCOFOR Project: http://www.joanneum.at/encofor.
38
Research Reports
88. "Bright Spots" in Uzbekistan, Reversing Land and Water Degradation While
Improving Livelihoods: Key Developments and Sustaining Ingredients for
Transition Economies of the former Soviet Union. Andrew Noble, Mohammed ul
Hassan and Jusipbek Kazbekov. 2005.
89. Planning for Environmental Water Allocations: An Example of Hydrology-based
Assessment in the East Rapti River, Nepal. V. U. Smakhtin and R. L. Shilpakar.
2005.
90. Working Wetlands: Classifying Wetland Potential for Agriculture. Matthew P.
McCartney, Mutsa Masiyandima and Helen A. Houghton-Carr. 2005.
91. When “Conservation” Leads to Land Degradation: Lessons from Ban Lak Sip,
Laos. Guillaume Lestrelin, Mark Giordano and Bounmy Keohavong. 2005.
92. How Pro-Poor are Participatory Watershed Management Projects?—An Indian
Case Study. Mathew Kurian and Ton Dietz. 2005.
93. Adoption and Impacts of Microirrigation Technologies: Empirical Results from
Selected Localities of Maharashtra and Gujarat States of India. Regassa E.
Namara, Bhawana Upadhyay and R. K. Nagar. 2005.
94. Balancing Irrigation and Hydropower: A Case Study from Southern Sri Lanka.
François Molle, Priyantha Jayakody, Ranjith Ariyaratne and H.S. Somatilake.
2005.
95. Irrigation and Water Policies in the Mekong Region: Current Discourses and
Practices. François Molle. 2005.
96. Locating the Poor: Spatially Disaggregated Poverty Maps for Sri Lanka. Upali A.
Amarasinghe, Madar Samad and Markandu Anputhas. 2006.
97. Strategies to Mitigate Secondary Salinization in the Indus Basin of Pakistan: A
Selective Review. M. Aslam and S. A. Prathapar. (not published yet) 2006.
98. Multiple-Use Water Services to Advance the Millennium Development Goals.
Barbara van Koppen, Patrick Moriarty and Eline Boelee. 2006.
99. Irrigation and Schistosomiasis in Africa: Ecological Aspects. Eline Boelee and
Henry Madsen. 2006.
100. The Reliability Improvement in Irrigation Services: Application of Rotational Water
Distribution to Tertiary Canals in Central Asia. Iskandar Abdullaev, Mehmood Ul
Hassan, Herath Manthrithilake and Murat Yakubov. 2006.
101. Carbon, Land and Water: A Global Analysis of the Hydrologic Dimensions of
Climate Change Mitigation through Afforestation/Reforestation. Robert J. Zomer,
Antonio Trabucco, Oliver van Straaten and Deborah A. Bossio. 2006.
RESEARCH
R E P O R T
Carbon, Land and Water:
101 A Global Analysis of the
Hydrologic Dimensions of Climate 
Change Mitigation through 
Afforestation/Reforestation 
Robert J. Zomer, Antonio Trabucco, Oliver van Straaten
and Deborah A. Bossio
Postal Address
P O Box 2075
Colombo
Sri Lanka
Location
127, Sunil Mawatha
Pelawatta
Battaramulla
Sri Lanka
Telephone
+94-11-2787404
Fax
+94-11-2786854
E-mail
iwmi@cgiar.org
Website
http://www.iwmi.org
I n t e r n a t i o n a l
SM
I n t e r n a t i o n a l Water  Management
Water  Management ISSN 1026-0862 IWMI is a Future Harvest Center
IWMI is a Future Harvest Center
I n s t i t u t e supported by the CGIAR ISBN 92-9090-641-3 I n s t i t u t e supported by the CGIAR
ISBN 978-92-9090-641-4