You are on page 1of 53




This publication was produced for review by the United States Agency for
International Development. It was prepared by Tetra Tech.

This publication was produced for review by the United States Agency for International Development by Tetra Tech ARD,
through a Task Order under the Prosperity, Livelihoods, and Conserving Ecosystems (PLACE) Indefinite Quantity
Contract Core Task Order (USAID Contract No. EPP-I-00-06-00008-00, Order Number AID-OAA-TO-11-00022).

Tetra Tech
159 Bank Street, Suite 300
Burlington, Vermont 05401 USA
Telephone: (802) 658-3890
Fax: (802) 658-4247

Tetra Tech Contacts:
Ian Deshmukh, Senior Technical Advisor/Manager

Forest Carbon, Markets and Communities (FCMC) Program
1611 North Kent Street
Suite 805
Arlington, Virginia 22209 USA
Telephone: (703) 592-6388
Fax: (866) 795-6462

Stephen Kelleher, Chief of Party

Olaf Zerbock, USAID Contracting Officer’s Representative

• Jennifer Hewson, Conservation International
• Marc Steininger, FCMC & Conservation International
• Stelios Pesmajoglou, Greenhouse Gas Management Institute

Contributing authors:
• Angel Parra, Consultant; GHG inventory & LULUCF sector expert
• Gordon Smith, Greenhouse Gas Management Institute
• David Shoch, TerraCarbon, LLC
• John Musinsky, National Ecological Observatory Network
• Fred Stolle, World Resources Institute
• Kemen Austin, World Resources Institute
• Irene Angeletti, Greenhouse Gas Management Institute

The US Agency for International Development (USAID) has launched the Forest Carbon, Markets and Communities
(FCMC) Program to provide its missions, partner governments, local and international stakeholders with assistance in
developing and implementing REDD+ initiatives. FCMC services include analysis, evaluation, tools and guidance for
program design support; training materials; and meeting and workshop development and facilitation that support US
Government contributions to international REDD+ architecture.

Please cite this report as:
Hewson, J., M.K. Steininger and S. Pesmajoglou, eds. 2014. REDD+ Measurement, Reporting and Verification (MRV) Manual,
Version 2.0. USAID-supported Forest Carbon, Markets and Communities Program. Washington, DC, USA.





The author’s views expressed in this publication do not necessarily reflect the
views of the United States Agency for International Development or the United
States Government.


TABLE OF CONTENTS....................................................................................... III
ACRONYMS AND ABBREVIATIONS ...................................................................... V
ACKNOWLEDGEMENTS ................................................................................... VIII
1.0 INTRODUCTION....................................................................................... 1
1.1 PURPOSE, SCOPE AND STRUCTURE ...................................................................................... 1
1.2 BACKGROUND ............................................................................................................................... 4
1.3 REFERENCES...................................................................................................................................... 8
2.0 INSTITUTIONAL ARRANGEMENTS.............................................................. 9
2.1 INTRODUCTION ............................................................................................................................ 9
2.2 ELEMENTS OF A MRV SYSTEM FOR REDD+ .......................................................................13
2.4 STEPS IN ESTABLISHING INSTITUTIONAL ARRANGEMENTS.....................................19
2.5 EXAMPLES ........................................................................................................................................22
2.6 EPA NATIONAL SYSTEM TEMPLATES ...................................................................................26
2.7 REFERENCES....................................................................................................................................30
3.1 INTRODUCTION ..........................................................................................................................31
3.2 IPCC GUIDANCE...........................................................................................................................34
3.3 INVENTORY AND REPORTING STEPS .................................................................................38
3.4 DEFINITIONS OF CARBON POOLS AND LAND USES ..................................................41
3.6 REFERENCES....................................................................................................................................49
4.0 FIELD-BASED INVENTORIES .................................................................... 50
4.1 INTRODUCTION ..........................................................................................................................50
4.2 CARBON POOLS AND THEIR MEASUREMENT ................................................................54
4.3 CONCEPTS AND CONSIDERATIONS IN INVENTORY DESIGN ................................ 57
4.4 THE FOREST CARBON INVENTORY TEAM .......................................................................66
4.5 FIELD WORK AND ANALYSIS .................................................................................................67
4.6 CALCULATING CARBON STOCKS FROM FIELD DATA ...............................................69


4.7 DATA CHECKING ........................................................................................................................75
4.8 CONSOLIDATING INVENTORY DATASETS ......................................................................76
4.9 THE GAIN-LOSS METHOD ........................................................................................................78
4.10 REFERENCES .................................................................................................................................79
4.11 SELECTED RESOURCES ............................................................................................................81
5.0 REMOTE SENSING OF LAND COVER CHANGE ........................................... 82
5.1 INTRODUCTION ..........................................................................................................................82
5.2 LAND USES AND CATEGORIES IN THE UNFCCC ..........................................................84
5.3 OVERALL STEPS AND NEEDS...................................................................................................90
5.4 REMOTE SENSING OVERVIEW ................................................................................................93
5.5 EMERGING AREAS OF RESEARCH....................................................................................... 109
5.6 REFERENCES................................................................................................................................. 114
5.8 SELECTED RESOURCES............................................................................................................ 121
6.1 INTRODUCTION ....................................................................................................................... 125
6.2 REPORTING.................................................................................................................................. 127
6.3 VERIFICATION ............................................................................................................................ 141
6.4 REFERENCES................................................................................................................................. 151
7.0 THEMATIC REVIEWS ............................................................................. 152
7.1 HISTORY OF REDD+ UNDER THE UNFCCC .................................................................. 152
7.2 COMMUNITY-BASED MONITORING ................................................................................ 165
7.3 NEAR-REAL TIME MONITORING AND ALERT SYSTEMS ........................................... 181


Markets and Communities Program FCPF Forest Carbon Partnership Facility REDD+ MEASUREMENT. REPORTING AND VERIFICATION (MRV) MANUAL. ACRONYMS AND ABBREVIATIONS ACR American Carbon Registry AD Activity Data AFOLU Agriculture. VERSION 2. Forestry and Other Land Use AGB Aboveground biomass BCEFs Biomass conversion and expansion factors BRDF Bi-directional reflectance distribution function BURs Biennial Update Reports CH4 Methane CI Conservation International CMP Conference of the Parties serving as the Meeting of the Parties to the Kyoto Protocol CO2 Carbon dioxide COP Conference of the Parties CV Coefficient of Variation DBH Diameter at Breast Height DEM Digital Elevation Model DTs Decision Trees EFDB Emissions Factor Database EFs Emissions Factors EM Electromagnetic EOS Earth Observation System EPA Environmental Protection Agency FAO Food and Agriculture Organization FAS Fire Alert System FCMC Forest Carbon.0 v .

Meteorology and Environmental Studies ILUA Integrated Land Use Assessment INPE Brazilian National Space Research Institute IPCC Intergovernmental Panel on Climate Change KCA Key Category Analysis LDCM Landsat Data Continuity Mission LEDS Low Emission Development Strategies LiDAR Light Detection and Ranging LUC Land-use Change MADS Colombian Ministry for Sustainable Development MCT Brazilian Ministry of Science.FIRMS Fire Information and Resource Management System FREL Forest Reference Emission Level FRL Forest Reference Level FSI Forest Survey of India FUNCATE Foundation of Space Science. Technology and Innovation MMU Minimum-mapping unit MRV Measurement. Reporting and Verification N20 Nitrogen oxide NAMA Nationally Appropriate Mitigation Strategies NASA National Aeronautics and Space Agency REDD+ MEASUREMENT. VERSION 2. Land-use Change and Forestry GPS Global Positioning System IDEAM Colombian Institute for Hydrology. REPORTING AND VERIFICATION (MRV) MANUAL. Applications and Technology GEF Global Environmental Facility GFIMS Global Fire Information Management System GFOI MGD Global Forest Observation Initiative Methods and Guidance Documentation GFW Global Forest Watch GHG Greenhouse gas GHGMI Greenhouse Gas Management Institute GIS Geographic Information System GLAS Geoscience Laser Altimeter System GOFC-GOLD Global Observation of Forest and Land Cover Dynamics GPG-LULUCF Good Practice Guidance for Land Use.0 vi .

NCs National Communications NFMS National Forest Monitoring System NGGIP National Greenhouse Gas Inventories Program NGO Non-governmental organization NNs Neural Networks NRT Near-real Time PCA Principal components analysis PRODES Projeto De Estimativa De Desflorestamento da Amazoni (Brazilian Amazon deforestation monitoring program) QA/QC Quality Assurance and Quality Control QUICC Quarterly Indicator of Cover Change RADAR Radio Detection and Ranging REDD+ Reducing emissions from deforestation and forest degradation. VERSION 2. completeness. SBSTA Subsidiary Body on Scientific and Technical Advice SES Social and Environmental Soundness SINA Colombian National Environmental System SLR Side Looking RADAR SRTM Shuttle Radar Topography Mission TACCC IPCC principles of transparency. accuracy. comparability. and consistency TOA Top-of-atmosphere UMD University of Maryland UNDP United Nations Development Programme UNEP United Nations Environment Programme UNFCCC United Nations Framework Convention on Climate Change USAID United States Agency for International Development USGS United States Geological Survey VCS Verified Carbon Standard WGs Working Groups WMO World Meteorological Organization WRI World Resources Institute REDD+ MEASUREMENT. plus the role of conservation. REPORTING AND VERIFICATION (MRV) MANUAL.0 vii . sustainable forest management and enhancement of forest carbon stocks.

Deborah Lawrence of the University of Virginia. including Colin Silver. The authors also thank those who have made graphics available for use in this Manual. John Rogan of Boston University. SilvaCarbon Consultant. REDD+ MEASUREMENT. Ronald McRoberts. Christine Dragisic of the US State Department. the International Panel on Climate Change and the Landsat program. Rishi Das. Axel Penndorf of BlackBridge. Michael Gillenwater of the Greenhouse Gas Management Institute. Chris Potter of the US National Aeronautics and Space Administration’s Ames Research Center. Frank Martin Seifert of the European Space Agency's Centre for Earth Observation ESRIN. Karyn Tabor. Mario Chacon and Johnson Cerda of Conservation International. Andrew Lister. Additional figures are from the websites of the United Nations Framework Convention on Climate Change. and Leif Kindberg of FCMC. Carly Green and Jim Penman of the Global Forest Observations Initiative. and Jamie Eaton of TerraCarbon LLC. Asim Banskota of the University of Minnesota.0 viii . VERSION 2. REPORTING AND VERIFICATION (MRV) MANUAL. Megan McGroddy. including Ned Horning of the American Museum of Natural History. and Charles Scott of the US Forest Service. Maggie Roth. Brice Mora of the Global Observation of Forest Cover and Land Dynamics. and members of the USAID Climate Change office.ACKNOWLEDGEMENTS The authors thank the various colleagues in our organizations who have commented on the text.

Inventory and Reporting Steps. REDD+ MRV MANUAL: CHAPTER 5.3.0 REMOTE SENSING OF LAND COVER CHANGE Authors: Marc Steininger and Jennifer H ewson 5. This chapter is relevant to the activities highlighted on the following page.1 INTRODUCTION This chapter focuses on the application of remote sensing-based approaches to forest cover and change monitoring. of this Manual outlines the sequence of steps required for generating a national greenhouse gas (GHG) inventory. Section 3.0 – REMOTE SENSING OF LAND COVER CHANGE 82 . 5.

STEP 0: Establish Institutional Arrangements. and worksheets where appropriate. if using the gain-loss method. AD represents the extent over which a human activity occurs. Document and archive information used to produce the national emissions and removals estimates following specific instructions under each land-use category. including expert peer review of the emission estimates following specific guidance under each land-use category. STEP 3: Design a forest carbon inventory to generate Emissions Factors (EFs). for the time period required. appropriate to the tier level identified. STEP 6: Report emissions and removals estimates. STEP 5: Quantify emissions and removals. Emissions and removals estimates represent the product of the AD by the associated EFs. using the reporting tables. STEP 7: Perform verification and implement quality control checks.0 – REMOTE SENSING OF LAND COVER CHANGE 83 . estimating the uncertainty in each estimate. REDD+ MRV MANUAL: CHAPTER 5. EFs represent coefficients that quantify the emissions/removals per unit area. STEP 2: Conduct key category analysis (KCA) for the relevant categories. ensuring that the requirements in terms of emission and removal factors are met. assess significant non-CO2 gases and carbon pools and prioritize such pools in terms of methodological choice. Within the categories designated as key. pool or non-CO2 gas. through stratification and other methods. carbon pool and non-CO2 source. for representing areas in the GPG-LULUCF. STEP 1: Estimate the land areas in each land-use category. STEP 4: Generate Activity Data (AD).

other valuable resources include the Global Observation of Forest and Land Cover Dynamics (GOFC-GOLD) Sourcebook (GOFC-GOLD. Additional forest strata could be of interest for national management and planning purposes and such stratification activities can be facilitated through the use of remotely-sensed data. Wetlands. Grassland. Settlements. particularly in the stratification of forests for field sampling. This chapter emphasizes optical satellite remote sensing of deforestation. Cropland. while the Guidance from 2006 uses Stock-Difference (IPCC. For land use.2 LAND USES AND CATEGORIES IN THE UNFCCC LULUCF. 2006). and Other Land (IPCC 2006. Optical satellite remote sensing is the most heavily used type of remote sensing for this application. within the context of the United Nations Framework Convention on Climate Change (UNFCCC). where estimates of changes between uses is a specific input to emissions estimates. The Intergovernmental Panel on Climate Change (IPCC) Good Practice Guidance for Land Use. and • An overview of emerging areas of remote sensing-based research for forest monitoring..In the context of the activities highlighted above. especially satellite-based approaches. This is needed to define the national forest area at the beginning of a reporting period and within which carbon stocks and forest changes will be monitored. This chapter discusses remote sensing of land-use change in the context of the gain-loss method. 2013). the IPCC recognizes two methods to estimate carbon emissions: the stock change method 23 and the gain-loss method (IPCC. and deforestation represents the largest source of GHG) emissions from the land-use sector in most tropical- forest countries. • Overall steps and needs for consideration in developing a satellite-based forest monitoring system. Finally. Land-Use Change and Forestry (GPG-LULUCF) is a key resource for countries. REDD+ MRV MANUAL: CHAPTER 5. this chapter discusses: • The context of land uses within the United Nations Framework Convention on Climate Change (UNFCCC). 4. see Chapter 2). 2003) uses the term Stock-Change. the Global Forest Observations Initiative (GFOI) Methods and Guidance Document (MGD) (GFOI. The information in this chapter summarizes remote sensing issues for a NFMS. Sub- categories can be defined within a category to more precisely define changes and emission sources. these six broad uses are called Categories. 2013).0 – REMOTE SENSING OF LAND COVER CHANGE 84 . However. Another important use of remote sensing in a National Forest Monitoring System (NFMS) is to produce a forest benchmark map. and the Forestry and Forest Products Research Institute’s REDD-plus CookBook (Hirata et al. However. 2012). and spectral variations due to very different canopy structures. as noted in Chapter 3. As mentioned in Chapter 1. refers to land-use change or persistence among the six broad uses defined by the IPCC: Forest Land. 23 The GPG-LULUCF (IPCC. or persistence within. 2003). even though the 2006 Guidelines are more up-to-date and use the latter. Links to additional resources for training on remote sensing are provided in Section 5. There has been no decision for non-Annex regarding use of the 2006 Guidelines. Possible types of land-use change among.8. • A brief review of remote sensing for forest monitoring. and thus in this Manual we use the former term throughout. 5. the remote sensing techniques and issues discussed in this chapter are also relevant to application in the Stock-Change method. remotely- sensed data represent a key input to a stratification of forest types and can be used to characterize the seasonality of leaf cover. provides the most practical option for monitoring land cover change over large areas. Stratification should seek to identify forest types with potentially significantly different levels of biomass to assist field sampling strategies (see Chapter 4). it provides limited information on specific approaches to remote sensing of land use. inundation. Vol. Remote sensing.

Case studies of countries that have used different approaches are provided in Annex 2A. thus enabling Approach 2. thus providing the data on differences in the carbon stocks before and after the cover change has occurred. While countries must report on land use. however.g. some types of land-use categories (e. or regions (e. the coefficients quantifying the emissions/removals per unit area. can usually be inferred based on local context and a general knowledge of the area. forest degradation). existed for the entire region. and Other Land Use (AFOLU) in Volume 4 (IPCC 2006. yet on a non-mapped manner Approach 3: Tracking of land-use conversion on a mapped manner Approach 3 is most informative and applicable to a mechanism for reducing emissions from deforestation and forest degradation. are used to estimate AD for each Category: Box 5. Forest definition A fundamental step in the development of a MRV system is the national definition of forest. The IPCC (2006) describes three overall approaches. As illustrated in Chapter 3 (Figure 3. Land use. For example. See Chapter 3). thus enabling Approach 1. plus the role of conservation. the creation of a multi-temporal map of change in forest cover as well as some sub-categories enabled Approach 3 for those categories.. 5. for the representation of land use (see Chapter 3). Multiplying the AD. However. listed in Box 5. not to be confused with tiers. it is acceptable to use a mix of the three approaches among regions or categories in a country. Data existed on land-cover change. Forestry. may be more effectively monitored through airborne or field-based data collection approaches..1 Definition of national forest and other classes The 2006 IPCC Guidelines for National Greenhouse Gas Inventories consolidated both LULUCF and Agriculture into the Agriculture.1: Approaches Approach 1: Net area of each land use reported at different time period. sustainable forest management and enhancement of forest carbon stocks in developing countries (REDD+). over a given period of time. In Australia.1. yet they are constrained by certain criteria. yet no tracking of specific conversions among them Approach 2: Tracking of land-use conversions.2). Throughout this chapter. documenting transformations between natural grasslands to pasture and cropland. are data on the area of a Category that potentially results in GHG emissions or removals. AD are combined with Emissions Factors (EFs). EFs are coefficients that quantify the emissions/removals per unit area.2. unless noted otherwise. as it is based on relationships between observed spectra in the images and the structural characteristics of the soil and vegetation covering land. The costs associated with these approaches could be significant and thus necessitate a sampling-based approach.Activity data (AD). These approaches. provides the estimated associated GHG emissions for each category. mountainous areas). While satellite-based remote sensing is a valuable tool for monitoring several parameters of land use. The definition REDD+ MRV MANUAL: CHAPTER 5. documenting the area of each land use over time and with full coverage. It is important to consider the characteristics of land-use parameters that will be monitored and the cost implications of a full-coverage mapping versus a sampling-based method.g. definitions have been adapted from the 2006 IPCC Guidelines and the GPG-LULUCF 2003. Countries have some flexibility in developing their forest definition.1 of the GPG-LULUCF.0 – REMOTE SENSING OF LAND COVER CHANGE 85 . existing data available for the Argentine Pampas were sufficient for either Approach 1 or 2. the extent over which a human activity occurs. by the EFs. satellite monitoring is more suited to detecting land cover. or the extent over which a human activity occurs. Agricultural census data.

Therefore.” they are part of an agricultural cycle with a defined temporal period. Tree-crown cover is not the same as leaf cover. In this case. if agricultural fallows 24 Examples of national forest definitions are available at http://cdm. The development of the forest benchmark map should reflect the national forest definition. there are carbon stock losses in the ‘forest remaining forest’ class.unfccc. and estimation of changes within the map extent. stratification and change estimation The development of forest mapping for REDD+ activities can be considered as a trio of components: creation of a forest benchmark map. For example. For example. they belong to a non-forest category. The physical criteria for forest. as tree crown cover is defined by the periphery of the crown.2. Considering agricultural fallows as non-forest greatly facilitates reporting on deforestation and associated GHG emissions. but has undergone degradation. then the site remains forest. since a country would not be required to estimate rates of the appearance of new fallows and their re-clearance when reporting changes in forest area 24.html REDD+ MRV MANUAL: CHAPTER 5. and are expected to be re-cleared after that while an urban park or agricultural fallow may meet the physical criteria of forest. Box 5. and the range countries can select for their definition. and Minimum patch size of 0. it would still be defined as forest despite the structural change.e. as much of their agricultural land is in some stage of fallow. It is also preferable to conduct these three activities one at a time. forest concession subjected to some selective-logging cycle). aerial photos. Agricultural fallow is a particularly important example for many tropical countries. The forest benchmark map should be created first. If the site is still under a forest use (i. A site is defined as forest if it meets the above criteria and if its main use is assumed to be forest-related. non-forest uses).05 ha to 1 ha. attempting to do all in a single process can lead to slow and overly-complicated processes. According to the IPCC Report Definitions and Methodological Options to Inventory Emissions from Direct Human- induced Degradation of Forests and Devegetation of Other Vegetation Types.2. as this defines the geographical extent for conducting the other two activities. the use of the land must also change. regrowing “forests.4 of the Kyoto Protocol” (IPCC 2003)..e. However. are included below in Box 5.0 – REMOTE SENSING OF LAND COVER CHANGE 86 . another site that has been logged and did cross below this threshold could be classified as deforestation. For example. while leaf cover is the proportion of leaf versus canopy-gaps. thus. Minimum tree-crown cover of 10 percent to 30 percent. these areas have urban and agricultural uses (i. very high resolution imagery and expert knowledge as possible to facilitate remote sensing imagery interpretation. they are part of a non-forest use. If the tree cover was not reduced enough to pass below the threshold of the forest definition. forest degradation could be defined as “a direct human-induced long-term loss (persisting for X years or more) of at least Y percent of forest carbon stocks [and forest values] since time T and not qualifying as deforestation or an elected activity under Article 3. selective logging may occur in a site defined as forest. Forest benchmark. While in terms of structure these are young. stratification of forests within the map extent. and use as much field information.. The GFOI MGD provides a further discussion of forest degradation in Chapter 2.must be developed based on both the physical structure of the present and potential vegetation as well as how the land is used. This will likely necessitate a subclass for ‘intact to degraded forest. in addition to the change in physical structure.2: Definition of forest criteria Potential to reach a minimum canopy height at maturity of 2m to 5m.’ and this subclass should be sampled to estimate carbon stock change. Conversely.2.

The creation of a forest benchmark map facilitates the estimation of change within forests.are excluded from the national definition. or Section 3. Section 3. Therefore. as discussed in Chapter 4. forest clear-cutting (removal of all trees) should be estimated. it will eventually need to be updated to define the extent of forest at the beginning of the implementation of REDD+ activities. smaller than a few hectares.e. and these can be re-combined later if field data indicate otherwise.g. This could be the date of initiation of REDD+ activities.2 of the GOFC-GOLD SourceBook (GOFC-GOLD 2013) provide useful overview information on the range of approaches and methodologies available. Stratification based on satellite imagery can benefit from the inclusion of seasonal information.. and palm forests. This is a critical step for analysts to correctly train and assess classifications of forest for the benchmark period. etc. Section 5. While the inclusion of the smallest clearances. of this Manual.2 of the GFOI MGD (GFOI 2013). Other characteristics often discernable from satellite data are major differences in canopy structures. A decision should also be made on the minimum size of clearance to be estimated. It is important that a final benchmark map is communicated to stakeholders with any concerns addressed. such as those of inundated forests.5.2 of the GFOI MGD (GFOI 2013). liana forests.2. The benchmark map should be based on satellite data from dates closest to the start date of a monitoring period. as this may result in the inclusion of forests that were cleared just before the start date. satellite sensor problems. as this is the main emissions source for many countries.2 of the GOFC-GOLD SourceBook (GOFC-GOLD 2013).0 – REMOTE SENSING OF LAND COVER CHANGE 87 . as this sets an important precedent for where REDD+ activities may or may not be implemented. At a minimum. aerial photos or very high resolution imagery could be obtained over sample areas to allow analysts to train themselves in the interpretation of fallow versus mature forest in different parts of a country. e. and Step 5. see Chapter 2). such as data on climate. it will necessitate several iterations of image analysis followed by reviews from experts and stakeholders. It could also be the start date of a historical analysis for use in estimating a Forest Reference Emissions Level or Forest Reference Level (FREL/FRL. Many approaches exist for estimating change within forests. The creation of a forest benchmark map will require a careful and iterative process. Stratification of forests within the benchmark area should be conducted in coordination with the field-based inventory team (See Chapter 4) and with the guidance of an expert statistician. such as open woodland. Other characteristics that analysts should pay special attention to include gradations to lower-stature vegetation that fall below a threshold in the national forest definition. such as monthly image composites from MODIS. If such field data do not exist. such as montane shrubs. although it is recommended as a method to reduce both field-survey costs and carbon-stock uncertainties. thus. Step 5. Once the forest benchmark map is finalized. Stratification is not required. and thus it defines the extent within which REDD+ crediting may be possible and where monitoring is required. In the latter case. analysts will need to identify an appropriate methodology for analyzing multiple images. Section 5. A key consideration is the type of change to estimate. such as sub-national governments and REDD+ project implementers. Satellite images for an exact date can be difficult to acquire due to cloud cover or gaps in data collection. can REDD+ MRV MANUAL: CHAPTER 5. or from very high resolution imagery. of this Manual.. gaps can occur as a result of cloud cover.2.4. where more deciduous forests may have significantly lower biomass levels. image dates used for the benchmark map may vary. resulting in fewer total strata. A forest benchmark map should have as few data gaps as possible. Generally. strata can be created based on expert opinions of forest types that should have different biomass levels. any areas defined as unmanaged forest (see Chapter 3) can be omitted in order to provide a final definition of where REDD+ can be implemented and where monitoring must be conducted. and elevation. Seasonal information could be obtained from coarser resolution data. Different potential strata can be assessed by merging with any existing data on carbon stocks. soils. This may necessitate the acquisition of multiple images for the same area and. i. A conservative approach is to avoid using images from before the start date. and those strata with little difference in stocks may be combined. if acquired for particular leaf-on and leaf-off months. Even the creation of a few broad classes is very useful for these purposes.5. for example. or open-canopy vegetation.4. See. Stratification can also make use of ancillary GIS data.

1 and 5.potentially yield a more correct and higher rate. may be considered. However. For example.1 presents an example of a land-cover change matrix with three broad categories (forest.0 – REMOTE SENSING OF LAND COVER CHANGE 88 . following the creation of the forest benchmark map. For example. Again. a country will conduct a stratification for a national forest inventory. as it can be assumed there are no transitions among these naturally- occurring vegetation types over the required reporting periods. transitions among different post- deforestation land uses do occur over short time periods. a country could first produce an estimate of forest clear-cutting using one approach. depending on the season and stage of crop development at the acquisition times of available images. not all of these must be estimated in the same process. Tables 5.. Again. the time and resources involved may be costly. managed grassland. and countries should consider the trade-off that usually exists between thematic precision of a land-use change study and the accuracy of the change estimates (e. such as sampling of higher-resolution data or even targeted field surveys. as well as any important transitions among those. an initial step could be to produce a forest benchmark map.g. Other classes It may be important to further stratify the six broad use classes into sub-classes where carbon stocks vary significantly. In contrast.2 illustrate this below. For forests. for degradation. cropland. A second step could be to produce a map of a single. Including sub-classes may provide data that are very useful for REDD+ national strategies and management policies. 1999. By combining the two. deforestation can be attributed to different forest sub-classes. A step-wise process may be worth exploring that uses different methods and levels of detail for different monitoring aspects. broad deforestation class that occurs anywhere within this benchmark. and many different opinions within the research community. Table 5. Mather. with forest sub-classes that have significantly different carbon stocks. For example. degraded forest. plantations and fallows may be difficult to distinguish. Conversely. REDD+ MRV MANUAL: CHAPTER 5. countries should have strong justification for including sub-classes since this will require methodologies that consume more resources. especially when one cannot be very selective about the season of the imagery used for analysis because of frequent cloud cover. could lead to greater classification error. This is an area where there are many options. such as wildfire.2 presents an example with greater thematic precision where the broad categories have been divided into sub-classes. Other forms of degradation. nonforest). and may not reflect a significant source of emissions. This can be performed one time. and then use a different approach. Combining approaches like these can provide all of the necessary estimates to complete a full land-cover change matrix. Degradation from selective logging is more difficult to detect and often only detectable for more intensive forms and with data obtained soon after the logging event. The expected increase in the accuracy and overall usefulness of emissions estimates should be clearly assessed. 2000). This should be assessed as part of both the national Key Category Analysis (KCA) discussed in Chapter 3 and the forest stratification process discussed in Chapter 4. A third step could be to use samples of airborne or other very-high resolution data to estimate the proportions of sub-classes of non-forest following deforestation. Table 5. Foody. strong justification in terms of improvement of emissions estimates is needed to justify attempts to include transitions among these classes. Spectral distinction of these uses is often difficult. while not requiring a very difficult process of spectral classification of all transitions among sub-classes. Issues related to sub-classes are somewhat different for forest versus non-forest.

“Non-forest” includes all non-forest. 100.5 Table 5.970 25 2.000 Degraded 0. referred to in the IPCC as “Initial land-use class” and “Land use during reporting year. 4. gross deforestation plus forest degradation is 0.000 Non.1: Example of a land-use change matrix with few land-use classes and change categories.2 in the first row of (b)).9 101.3 100 T1 T1 Forest Forest Non.” Values in Sum T1 and Sum T2 are total area and percent change for each class. Values inside the matrix are areas and percent change for each category of persistence or change.0 100 forest forest Sum T2 9.940 40 20 10. In this example. T2 a) Forest Degraded Non-forest Sum T1 b) Forest Degraded Non.0 – REMOTE SENSING OF LAND COVER CHANGE 89 . “Forest” in this table is non-degraded forest only.4 0.045 % T2 99.2 100 Degraded 5 1.000 Forest 99.6 percent (adding values 0.010 4. a) T2 Degraded Degraded Lowland Montane Natural Lowland Montane Fallow Cropland Pasture Sum T1 Forest Forest Grassland Forest Forest Lowland Forest 7945 35 3 5 7 7995 Montane Forest 1995 5 2 3 2005 Degraded Lowland Forest 5 1500 2 6 12 1525 Degraded Montane Forest 470 1 4 475 T1 Natural Grassland 993 3 4 1000 Fallow 350 50 150 550 Cropland 200 700 100 1000 Pasture 50 1400 1450 Sum T2 7950 1995 1535 475 993 608 771 1673 REDD+ MRV MANUAL: CHAPTER 5. and in (b) are percentages.5 1. both naturally-occurring and anthropogenic. % Forest Forest forest T1 Forest 9.3 98. T1 and T2 are the first and second time periods.4 0.945 2.4 and 0. Values in (a) are in absolute units. such as hectares.7 98.000 4.

0 70. In this example.1 100 Montane Forest 99.4 96. the large areas of change from cropland to fallow (200) or pasture (100).2 99.5 98.1 0.1 100 Degraded Lowland Forest 0.2: Example of a land-use change matrix with more precise land-use classes and change categories. level of automation and analyst expertise. 5. A high degree of rotational land use is also indicated by. the majority of forest occurs in the lowlands.3 0.9 0.4 0. such as hectares and in (b) are in percent.3 100 Cropland 20. The 35. the majority of deforestation (to fallow.5 80.1 Table 5. pre-processing. non-degraded forest.3 98.1 0.8 99. “Forest” here means intact.4 0. REDD+ MRV MANUAL: CHAPTER 5. Values inside the matrix are areas and percent change for each type of category.1 27. b) T2 Degraded Degraded Lowland Montane Natural Lowland Montane Fallow Cropland Pasture % T1 Forest Forest Grassland Forest Forest Lowland Forest 99.8 135.4 0.0 100 Pasture 3.1 0. and pasture) and forest degradation also occurs in the lowlands.4 0.5 percent reduction in fallow indicates intensification of land use. Criteria include the type and resolution of satellite data and the degree to which a full coverage.7 99.” Values in Sum T2 are total area and percent change for each class.0 10.5 0. The appropriateness of different monitoring methodologies will need to be assessed.3 87.8 100 Degraded Montane Forest 98. or sampling-based.2 0. croplands. referred to in the IPCC as “Initial land-use class” and “Land use during reporting year. Cropland. T1 and T2 are the first and second time periods.4 100 Fallow 63. Each of these decisions is discussed below. Where automation is not possible. according to the national forest definition.1 percent increase in pasture indicates an increasing importance of this use. and Pasture represent non-forest classes. The 12.3 OVERALL STEPS AND NEEDS Figure 5.0 – REMOTE SENSING OF LAND COVER CHANGE 90 . either via a shortening of fallow cycles or an increase in permanent pasture. Fallow.6 100 % T2 99. it is important to consider how consistency will be achieved and what methods will be used to effectively combine data from different time periods.3 0.1 illustrates the key decisions a country should consider when developing an effective and efficient satellite-based forest monitoring system after identifying which categories and sub-categories to monitor and the scale of monitoring.0 0.2 0. classification algorithms. Natural Grassland. approach should be applied to monitoring different land cover classes. Values in (a) are in absolute units.8 100 T1 Natural Grassland 99.6 9. for example. including the types and availability of different satellite data.

a country must determine the geographical extent of managed land. whereas other dynamics. the process should extend to defining the types of land-cover changes that are major GHG contributors and to aligning these definitions with the land-cover change categories defined by the GPG-LULUCF. it is necessary to consider the appropriate scale and approach.Figure 5. some land-use dynamics may be very appropriate for satellite-based monitoring. For MRV. Finally. particularly some forms of degradation and post-deforestation land-use changes.0 – REMOTE SENSING OF LAND COVER CHANGE 91 . These latter land-use dynamics may require more costly data -collection REDD+ MRV MANUAL: CHAPTER 5. and thus where monitoring should be conducted (see Chapter 2). A KCA should be performed as part of the development of a REDD+ strategy within the national development-planning context. do change events occur in small patches of several hectares.1: Key considerations in the development of a NFMS 1) What categories are most important to monitor? A KCA involves identifying the major land-use-based sources of GHG emissions. or are they much larger? Different types of changes may also be most appropriately monitored with different sources of data. For example. For example. 2) What are the appropriate scales and/or sampling approaches for monitoring? Once the categories and classes to monitor have been assessed. may require airborne or field-based monitoring.

such as deciduous woodland. While older secondary forest can be difficult to distinguish from mature forest.processes and thus necessitate a sampling approach. Many differing views exist regarding optimal methods for monitoring land-use change and. By distinguishing mature forest versus fallow areas. or a case should be made that using a larger MMU does not significantly affect resulting area estimates. a country can minimize confusion between mature forest clearing and fallow cycles in later monitoring. again necessitating increased data and analysis demands. This will avoid either: (i) the use of methods with little justification based on in-country testing. focusing on the categories identified in the KCA. while using a larger MMU may not significantly affect area estimates for static areas of classes. and conduct assessments with national data. Some of the main questions to consider within the methodological options are: 1) What types of satellite data are most appropriate for monitoring the classes identified? 2) What type of classification approach should be used? 3) What types of pre-processing are needed for the particular method of image analysis being considered. For question 4. and often some type of filtering to a defined minimum-mapping unit (MMU). The MMU should be smaller than the minimum patch size included in the national forest definition. Question 7 should consider REDD+ MRV MANUAL: CHAPTER 5. A country should obtain opinions from a range of international experts. may require data from particular or multiple seasons within each year. However.” This is a fundamental requirement of the GPG-LULUCF. Question 6 includes merging of temporary sub-classes. what spatial resolution is needed and whether the data source has an appropriate archive and acquisition strategy. possibly combining information from multiple dates into a single multi-date product. Further.0 – REMOTE SENSING OF LAND COVER CHANGE 92 . a country should seek its own cadre of experts with strong fundamental backgrounds in remote sensing to fully access and understand the relevant literature and options. how can consistency and reliability be assured? 5) How should data from different time periods be combined to produce change estimates? 6) What post-classification processing steps should be applied? 7) What validation approach should be used. 3) What methodological aspects should be considered? A country should consider a range of methodological options. and what level of analyst expertise is required? 4) How much of the process can be automated. a country should seek to produce the most accurate estimates possible for key categories while using an approach that is “replicable. including data sources and sampling? Some of the most important considerations are whether to use optical data versus Radio Detection and Ranging (RADAR) data. and for those parts that are dependent on analyst interaction. Question 5 includes both the approach to processing the satellite imagery from multiple dates and the approach to estimating change rates from completed land-use change maps or sample estimates. therefore. although it is only vaguely defined in the context of satellite monitoring. and the great majority of fallow periods are shorter than 10 years. some vegetation types. either during creation of the forest benchmark map or during the stratification of forests in this map. or (ii) spending too much time investigating issues that can be well-informed by existing literature or are not significant to the potential accuracy of the final emissions estimates. estimates of change can be very sensitive to the MMU. most fallows younger than 10 years are distinguishable.

The portion of the sun’s energy across these spectral regions reflected by the land surface is often indicative of the structural and chemical characteristics of the surface features (Figure 5.4 REMOTE SENSING OVERVIEW This section provides a summary of remote sensing fundamentals. two broad types of remote sensing for monitoring land-cover exist: passive and active. Satellite-based remote sensing is most common because of the full. the instrument does not emit its own signal. the amount representing the strength of the signal. or entire countries. Passive remotely-sensed data are acquired by a sensor that passively receives energy originating from another source. Links to selected internet resources are provided in Appendix 5B. 5. sampling approach. Fundamental assumptions. thus enabling national monitoring for terrestrial-based applications. REDD+ MRV MANUAL: CHAPTER 5. Different spectral regions are represented by relatively narrow “spectral bands” (Figure 5.3). depending on the type of equipment. as such systems could be applied over large regions. The sun is the source for visible and shortwave-infrared spectral regions of the earth. Passive remote sensing The majority of remotely-sensed data used for monitoring land use is passive. the feature itself is the source for thermal-infrared regions (Figure 5.1 Types and characteristics of remote sensing data Remote sensing is the process of sensing energy emitted or reflected at some wavelength along the electromagnetic (EM) spectrum by an object. The amount and type of energy sensed is usually recorded in digital form. 5. are that different land-cover types can be distinguished based on this recorded information and that land use can be inferred from land cover. Numerous text books are also available on remote sensing of land-cover. as well as an independent team of analysts to interpret validation data and conduct error calculations.5. senses a relatively small portion of the total spectrum of energy emitted by the sun.4. “multi-spectral” images are produced. At the highest level. by combining images of energy measured in different spectral bands and assigning a separate color for display. this is the visible portion of the EM spectrum. as illustrated in Figure 5. repeated coverage offered by one or more satellite data sources. The human eye. though not always valid. Airborne remote sensing capacities are also of interest.2).4) and.0 – REMOTE SENSING OF LAND COVER CHANGE 93 . rather than being in direct contact with it. for example. and resources available.various sampling schemes and the availability of very high-resolution satellite or aerial observations. and the type representing the recording of the signal across a spectrum.

A single channel of a multi-spectral sensor is sensitive to energy only within a certain band. Both (a) and (b) represent the entire visible range of the electromagnetic spectrum. plus non-green vegetation and shadows caused by the geometry of terrain and vegetation. A sensor. with so many channels and bands would be considered hyper-spectral. passes again through the atmosphere and reaches a sensor on board a satellite. the bands cover a wide range of energy. Spectral bands are defined by a range of wavelengths. topography.0 – REMOTE SENSING OF LAND COVER CHANGE 94 . reflects off a surface. Shortwave energy is emitted by the sun. such as illustrated in (b). passes through the atmosphere.Figure 5. and in the example here they are divided by white lines. REDD+ MRV MANUAL: CHAPTER 5. Most types of land-cover are a mixture of these features. and a sensor with such bands would be considered a broad-band sensor.2: Optical satellite remote sensing.3: Generalized spectral curves of fundamental features in remote sensing of land-cover. and a sensor with channels along these bands would have a high spectral resolution. In (a). A) B) Figure 5. the bands are narrow. The signal detected is dependent not only on the reflectance properties of the surface but also on the sun angle. Figure 5. view angle and atmospheric properties.4: Example of spectral resolution. In (b).

and are extremely influenced by topography. or branches and tree trunks. while rarely used for distinguishing types of land-cover. Thermal energy is emitted by the land surface itself and. passive remote sensing systems also acquire data in the thermal region. Several recent satellites carry RADAR sensors that collect data in multiple bands and in different polarizations. 2011. Because of their sensitivity to the geometric properties of forests. leaf pigments. Active remote sensing In active remote sensing. and thus yielded limited potential for classification of land-cover types.Figure 5. Baccini et al. it facilitates the detection of clouds. 2012). and urban heat islands. active fires. RADAR and Light Detection and Ranging (LiDAR) are the most commonly used active remote sensing techniques for terrestrial applications. the microwave portion of the EM spectrum. for example.. and soil background. thus extending their utility for classification of land-cover types. RADAR data have potential for relating to forest biomass. RADAR data provide information related to the density of leaves in the canopy. RADAR information is primarily related to structural features at the wavelength scale of the energy being sensed. One major advantage of RADAR systems is their ability to penetrate clouds due to the longer wavelengths. RADAR is further discussed in Section 5. In forest environments. all RADAR sensors on board satellites collected measurements in only one wavelength band and one polarization. versus optical sensors which measure reflected energy that is largely a function of canopy architecture. In addition to the visible and near and shortwave-infrared regions. Such color-composite images aid visualization and interpretation of the land-cover.5: Image data combined from three sensor channels to produce a multi-spectral image.5. depending on the wavelength used. Until recently. RADAR data are also sensitive to canopy and soil moisture. represent values in the individual channels. an instrument sends out a signal at certain wavelengths and measures the return time and strength of the back-scattered signal.0 – REMOTE SENSING OF LAND COVER CHANGE 95 . as well as modeling various ecosystem processes and vegetation-climate interactions. The resulting images did not have the dimensionality that multi-spectral images have. Brightness levels. were used to produce two recent maps of global forest biomass (Saatchi et al. RADAR data. and data from other satellite sources.. REDD+ MRV MANUAL: CHAPTER 5. shown as grey tones.

. per transmitted pulse. thus. Nelson et al. usually between one and five. for example. However. Both forms of LiDAR have been shown to be useful for estimating forest biomass via comparison with field data and modeling.As with RADAR. 2002. The distance between the sensor and the target is then calculated from the elapsed time for the LiDAR signal to make a complete round trip. discrete waveform LiDAR (Lim et al. in contrast to RADAR. enabling the creation of images. LiDAR operates in the visible and near infra-red portions of the EM spectrum and. Applications of LiDAR in forestry have mainly focused on measurement of canopy height. and the horizontal and vertical distribution of vegetation.0 – REMOTE SENSING OF LAND COVER CHANGE 96 . . 2003. while discrete systems sample a discrete number of points. LiDAR systems are also generally classified into full-waveform LiDAR and discrete LiDAR systems (see Figure 5. Clark et al. Full Discrete Waveform Return LiDAR LiDAR Figure 5. some of which is scattered back to the sensor by the target. While some LiDAR instruments collect data only along sampling lines. these parameters can be used to model estimates of aboveground biomass (see. sub-canopy topography.. Lim et al.6: Example of full waveform vs.6).. Full-waveform systems record the entire waveform of a returning pulse. 2009).. 2004. 2003) REDD+ MRV MANUAL: CHAPTER 5. LiDAR instruments emit a pulse of energy. does not penetrate clouds. others have scanning abilities to collect data both along and across sampling lines. Lefsky et al.

The majority of LiDAR remote sensing to-date has been airborne-based. for example. GLAS provided full- form LiDAR information for linear tracks along the satellite path.. especially via sampling. The use of such data for land-cover monitoring yields an inherent bias in such derived estimates. Publically available data have spatial resolutions ranging from 0. Conversely. have defined orbits that dictate how frequently the satellite will return to view the same location on the earth and acquire a new image.0 – REMOTE SENSING OF LAND COVER CHANGE 97 . is becoming more practical.7 illustrates the archive history of the Landsat satellite series. Mexico utilizes Landsat in its MRV system and has also incorporated RapidEye. one LiDAR instrument. Other satellites. consideration must be given to: spatial and temporal resolution. Data with coarser spatial resolutions are used for global studies and are not generally suitable for land-cover monitoring.7m to 1km. the use of these data over large areas. as data extend back to 1972 for the multi-spectral scanner (MSS) and 1982 for Thematic Mapper. While this has historically limited the practicality of using such satellites in a monitoring capacity. However. Many satellites. persistent cloud cover often reduces the frequency of acquiring useable images.3 highlights two examples where very high-resolution imagery is being used extensively in the development of NFMSs. A 30m resolution observation. Data used for land-cover monitoring have resolutions ranging from 5m to 30m. as high resolution data continue to become available and more affordable. at least half the size of the scale of changes). and land Elevation Satellite (ICESAT) was satellite-based. some countries are investigating the utility of other sources of. especially in areas where degradation pervades. KOMPSTAT-2 or RESOURCESAT-1 are pointable. Temporal resolution refers to the frequency with which data are collected. it is preferable to work with a single source of data throughout a study period when possible. It uses RapidEye imagery from the dry and wet season to better differentiate between seasonal biomass coverage. with five MSI bands of 6. cost of procuring the imagery and image archive length. Resolution and other considerations In addition to the type and spectral characteristics of different images. has implemented a national-level. The Landsat series is the most common data source for monitoring land-cover change. For Guyana. including many of the high resolution sensors such as RapidEye. Cloud. the DMC constellation. Data archive length is another important consideration for developing historical analyses. data collection strategy. Quickbird. such as Landsat. The linear samples from GLAS were inputs to the two global biomass studies noted above. REDD+ MRV MANUAL: CHAPTER 5. Box 5. Examples of the use of high resolution imagery in NFMSs While many countries are generating forest benchmark maps and deforestation monitoring products using Landsat imagery. or pixel. as such data will not detect smaller-scale changes. To facilitate consistent monitoring and ease of logistics.e. IKONOS. However. a ground area of one hectare would be represented by 11 pixels at this resolution. annual monitoring activity using wall-to-wall RapidEye imagery. The re-visit time for Landsat is 16 days. Spatial resolution is important as the resolution must be fine enough to detect the changes of interest (i. However. CBERS HRC. higher resolution imagery for monitoring purposes.3. has traditionally only been used over small areas because of cost and frequency of availability. so the most that an area could be monitored is every 16 days.5m. even for periods as brief as the past decade. the Geoscience Laser Altimeter System (GLAS) on board the Ice. For example. less than 10m. such acquisitions occur for short periods and some require tasking. with a ground resolution of 70m. high resolution data. has proven particularly useful for assessing and addressing the impacts of forest degradation. GeoEye-1 & -2. for example. represents a ground area of 900 m2 (30m by 30m). While this can result in the same area being repeatedly imaged at much higher frequency. WorldView-2. Figure 5. meaning they can be tilted to view a location that is at an angle to their defined orbit. SPOT HRV series. Box 5. Though the ICESAT satellite is no longer operational. Guyana. RapidEye.

has a follow-on mission planned for 2019-2020. is listed in Section 5. the example of Sentinel-2 The European Space Agency (ESA) is responsible for the space component of the Copernicus programme. RapidEye.usgs. with five satellites in its current constellation. Landsat 8. The Landsat Data Continuity Mission (LDCM). and changes among them? • What is the spatial resolution of the data and how appropriate is it.4: Additional future data sources for consideration in the development of NFMSs. 20m and 60m. key data characteristics to consider are: • What geographical. In summary. phenological. thus providing another source of data for an extended period. full and open data policy adopted for the Copernicus programme. Landsat’s no-cost data policy allows flexibility in data use. A table of current and future satellite data options. Box 5. as well as cost policies. including characteristics. Sentinel-2A. Landsat 7 carries the ETM+ instrument but since 2003 has experienced data gaps due to mechanical failure. NIR and SWIR wavelengths and spatial resolutions of 10m. and bands within them. a 5-day revisit time will be possible.6. are being constructed.4). The Sentinel-2 series of multi-spectral imaging satellites.php Finally. ensures that future National Aeronautic and Space Agency (NASA) satellites will continue to provide a long-term data record. a satellite program’s future sensor launch and data acquisition strategy. Landsat 4 – 5 carried both the MSS & TM instruments. carrys the OLI and TIRS instruments. will provide additional data options for consideration in the development of NFMS. are important considerations when planning a monitoring program. and atmospheric (especially persistent cloud cover) conditions exist? • What are the spectral regions. with 2A scheduled for launch in 2015. meaning image pairs from multiple dates mostly overlap. Figure 5. The first satellite. for relative to the scale of the land- cover changes to monitor? REDD+ MRV MANUAL: CHAPTER 5. the Sentinel data products available to all users. roughly the same area is acquired each time the satellite returns to view the same location on the earth. will provide an additional source of imagery for consideration in the development of NFMS (see Box 5. will be launched between May-July 2015 and will offer a revisit time of 10 days. The upcoming launch of the Sentinel-2 series.7: Landsat archive timeline. including the successful launch of Landsat 8 in February 2013. under which a series of dedicated satellites. with 13 bands located in the VIS.0 – REMOTE SENSING OF LAND COVER CHANGE 98 .Further. where data are collected. Landsat 1 – 3 carried only the MSS instrument. the Sentinels. When the second unit (Sentinel-2B) is launched in 2016. From http://landsat. as the Landsat satellites have a defined orbit. Costs of other data sources are also trending downwards. Based on the free. The interoperability between these sensors and the Landsat missions will further enhance the revisit time and improve overall data availability. and how do these relate to the potential for distinguishing the land-cover types of interest.

This is to understand the geographical area represented and is applied when importing the image into a GIS or image- analysis format for processing. though useful. geometric registration may have errors up to 100s of meters. However.php REDD+ MRV MANUAL: CHAPTER 5. • What is the temporal resolution in terms of potential frequency of acquisition of non-cloudy observations compared to the desired frequency of monitoring? • What is the longevity of the image archive length – does this meet the historical mapping needs? • What are the cost implications of these data in terms of purchase and analysis? • What are the future satellite development and launch commitments? 5. Image pre-processing Geometric registration and co-registration Geometric registration is the process of mapping data in a geographical coordinate system. The summary of pre-processing. yielding a dataset with accuracies within 30m and. Pre-processing usually includes geometric registration and co-registration. 25 http://landsat. and post-processing steps below is based on optical data and approaches to classification. although images have been geometrically registered. Post-processing occurs after the image analysis step. eliminating the need for further geometric correction. atmospheric correction. As previously outlined. Automation is increasingly available for processing numerous images. thus. Co-registration is a standard. and image analysis is the process of generating a land-cover class for all parts of an image. simple process that takes a modest amount of time and involves the identification of one image to use as the base image to which the remaining images will be co-registered. and occasional data transformation. co- registration may still be necessary. analysis. and is dependent on both the incident and reflected directions. the calculation of change rates and error estimates are required. These data have already been geometrically corrected using precision ground control points and Shuttle Radar Topography Mission (SRTM) Digital Elevation Model (DEM) information. but traditional analyst- driven methods are also sufficient. Atmospheric correction The atmosphere has several effects on visible and infrared energy as it passes through the atmosphere from the sun to the land and back to a satellite or airborne sensor (Figure 5. is optional depending on the image analysis approach. using examples of Landsat data analysis. Co-registration among images should be reviewed and may require adjustments. BRDF defines how light is reflected from a surface. Finally.4. and enables the estimation of rates and patterns of land-cover change to be generated. Therefore. Data transformation. post-processing activities may include a number of steps. Atmospheric correction is frequently performed in combination with a bi-directional reflectance distribution function (BRDF) correction. The United States Geological Survey (USGS) has been reprocessing much of the Landsat archive. Atmospheric correction may be necessary depending on the image analysis approach that will be used. and post-processing Image pre-processing refers to any step that is applied to an image in preparation for the image analysis step. Therefore.0 – REMOTE SENSING OF LAND COVER CHANGE 99 .2 Image pre-processing.usgs.8). it does not mean that images of the same area acquired from different dates will overlay well enough to avoid errors in change estimates resulting from poor co-registration. resulting in the creation of a L1T precision and terrain corrected product 25.

local histogram matching could be applied to further reduce artifacts.. 26 http://landsat. “gap-filled” composite mosaic. developed by NASA. and requiring each atmospherically-corrected image to be classified separately and then combined. The images can then be combined to produce a single. in unitless values from zero to one. LOTRAN and 6S are the most common.Atmospherically-corrected images contain data representing surface reflectance.php#3b REDD+ MRV MANUAL: CHAPTER 5. LEDAPS also generates water. and use a single correction algorithm for the entire image. atmospheric artifacts may remain. While the corrections are applied to all the images and the resulting composite should therefore appear seamless. Most atmospheric correction algorithms are applied to satellite images prior to mapping.9). and these methods are highly dependent on careful atmospheric correction. 2008). One example. additional algorithms based on. These corrections can be applied to several partially-cloudy images of the same area. appearing as darker or brighter patches (Figure 5. Several programs exist to perform atmospheric corrections over entire images. In addition. and several tools have been created to facilitate their application. A “clear” atmosphere still causes scattering and absorption of the radiation as it is transmitted from the sun to the earth and back to the satellite. as opposed to the digital numbers of the raw image data. will be available with top-of-atmosphere (TOA) reflectance corrections applied 26.. all Landsat data.8: Atmospheric effects on optical data. These usually assume constant atmospheric conditions across an image. is the Landsat Ecosystem Disturbance Adaptive Processing System (LEDAPS) tool (Masek et al. a topography-dependent Rayleigh scattering correction and an aerosol optical thickness component based on Kaufman et al. including those from the LDCM. (1997) to generate a surface reflectance value for each pixel. Conversely. BRDF characterizes how an object illuminated by a source. such as the sun. cloud. Figure 5. Some semi-automated methods apply constant signatures over multiple images or image dates. cloud shadow and snow masks. Approaches to classification that involve the creation of sub-classes for each type of land use and change can yield accurate maps without atmospheric correction because sub-classes can account for different atmospheric conditions. The electromagnetic radiation source is the sun. 2001). Alternatively.0 – REMOTE SENSING OF LAND COVER CHANGE 100 . appears brighter or darker depending on the angle of the source and the angle at which it is viewed by a satellite sensor. methods that apply constant class signatures over images with variable atmospheric conditions should include atmospheric correction. although there is active research on accounting for variability within an image. beginning in summer 2013. ozone. and this radiation can be blocked or scattered by clouds in addition to being affected by a “clear” atmosphere.usgs. Sun and sensor view angles also impact the effects of the Performing atmospheric correction depends on the image analysis approach used (Song et al. for example. atmospheric pressure. LEDAPS uses information on water vapor.

but with histogram matching applied. The black lines are SLC-off data gaps. a common assignment for a “false-color composite”. some of the cloud gaps remain since they were also cloud in the second image. green and blue bands are assigned to the red.9: Example of atmospheric correction of Landsat data from San Martin. note the orange-tone artifacts that appear to the left of the remaining cloud gaps in the upper-left of the image. Note that while the linear gaps have been filled. but post-atmospheric correction and with a cloud/cloud-shadow mask applied using LEDAPS.0 – REMOTE SENSING OF LAND COVER CHANGE 101 . green and blue colors. green and blue colors in the display. B) the same image. the artifacts in E) are no longer visible. C) the near- infrared. allowing visual exploration of the infrared data. REDD+ MRV MANUAL: CHAPTER 5. A) an unstretched “true-color composite” where the red. F) mosaic of the same two images .A) B) C) D) E) F) Figure 5. D) the same. E) a mosaic of two atmospherically corrected images. but with no histogram matching between them applied. thus requiring additional images. but with a Gaussian stretch applied to the data histogram. Peru. middle-infrared and red bands assigned to the red.

The “Tassled Cap” transformation has been used extensively in the classification of vegetation types. at a minimum. Often. and water could be included. and this may suggest a need to merge or add additional sub-classes. SMA involves defining the spectral reflectance of each main feature representing the landscape under study and. yielding a classified image. Spectral mixture analysis (SMA) is another data transformation approach. SMA does not add to the information content. they are referred to as “spectral end members. as it explains the data in terms of physical features. for example. unidentified pixels. that a particular pixel represents an observation of a piece of land that is 30 percent sunlit leaf. SMA can be a useful approach in understanding the spectral data contained in the image data. various algorithms can be used to estimate the most likely class of the remaining. based on these. Shadow is also generally included. SMA utilizes estimated spectral reflectances of a set of “pure” features that aim to represent the observed surfaces. variances and co-variance matrix of the spectral data. and other types of data ordination. Maximum Likelihood. in increasing complexity: Parallelepiped. Figure 5. Soil theoretically could be split into multiple soil types with different reflectance properties. When applied in SMA. training data are modified and a new iteration of the classification is run. Classification Land-cover classification produces a thematic representation of land by categorizing pixels based on their spectral signatures. these include the means. field analysis. Minimum Distance. Like PCA. Based on these statistics. soil. Two broad types of classification exist: supervised and unsupervised. End members can also be defined by simply selecting the extreme pixels in the multi- dimensional data. prior to classification. Statistics of the pixel data within these areas are calculated and. These can be visualized as “fractional images” and used as inputs to classifications. Principal components analysis (PCA) is one example of a transformation technique involving ordination. these techniques may help to produce more accurate or efficient land-use classification. In vegetated lands. estimating the proportions of each component for each pixel. The level of statistical separability among the classes can be evaluated. These techniques alter the information to facilitate interpretation.” However.Data transformation Some analysis methods include data transformation techniques. defining the “spectral signature” of each class. as this is an important feature of most spectral images because of the geometry of the vegetated canopies. and woody vegetation or litter. such as various forms of ordination.0 – REMOTE SENSING OF LAND COVER CHANGE 102 . if image end members are used. or literature. output classifications are evaluated and. An output could estimate. these are termed “image end members. the resulting SMA analyses are relevant only to that image. In supervised classification. Depending on the classification approach used. REDD+ MRV MANUAL: CHAPTER 5. for example.10 illustrates the supervised classification approach. Definition of the pure features may be via laboratory analysis. based on conspicuous errors. these are sunlit leaf. the analyst identifies “training sites” and creates training data by delineating areas known to be of each class. and Mahalanobis Distance. 20 percent soil and 50 percent shadow.” since they are located at the outer ends of the multi-dimensional distribution of the spectral data. Some of the common algorithms in remote sensing software packages are.

Instead. These are the basis for class statistics used to classify the rest of the image. Supervised algorithms that explore the distribution of the data while still allowing the analyst to direct the process via training are increasingly used. Additional iterations are typically run to further split groups that overlap different land-cover types. no training process is applied. Spectral images (A) from two dates can be combined and viewed to more easily observe spectral changes (B) indicative of land-use change. from Liberia. The output of an unsupervised classification is then reviewed by an analyst. Many variants of NNs exist. Pal and Mather. the resulting DT is a set of rules that is applied to the rest of the pixels to produce a classified image.10: Example of a supervised classification of two dates of images in a single process.A) B) C) D) Figure 5. 1999. and available ancillary field or other data. Rogan et al. The final product (D) is often filtered to eliminate small and spurious errors. 2003. The final tree is often composed of hundreds of splits and terminal nodes representing the land-cover classes contained in the training data. and each group is labeled to a class based on the analyst’s visual interpretation of the spectral data. 2008b). 2003. the location of the pixels. Hansen et al.. Conversely. The split is one that optimizes accuracy at that stage in the development of the DT. an assumed advantage of the unsupervised approach is the algorithm evaluates the distribution of the data itself. though NNs can run more slowly than DTs..0 – REMOTE SENSING OF LAND COVER CHANGE 103 . The ISODATA algorithm is common in most software packages. In unsupervised classification... NNs attempt to mimic the human learning process to associate a class with the image data. Hansen et al. Numerous studies have used DTs to generate robust classification results in many regions (Friedl et al. Both DTs and NNs have become favored over REDD+ MRV MANUAL: CHAPTER 5. 2000. based on the data in the training sites identified by the analyst. DTs operate by iteratively seeking a binary split in the data in each of the bands. algorithms identify spectrally similar pixels and then assign them to a user-specified number of groups. Training sites can be drawn (C) based on field and aerial data as well as an analyst’s knowledge of an area and expertise in interpretation. Two such algorithms are Decision Trees (DTs) and Neural Networks (NNs). “Boosting” and “pruning” processes can be applied to DTs in order to improve the efficiency and reduce the number of final splits. An assumed advantage of supervised over unsupervised classification is that the analyst directs the process based on a priori knowledge of the area being classified.

in which multiple decision trees are constructed and an output class is selected based on the majority of votes from all the trees. Other approaches under exploration include learning classification techniques.. 2008a. 2010. rather than the pixel level. Lindquist et al. meaning the pixel is classified solely based on its spectral characteristics. Random Forests. or further from.. which can be a problem with DTs. but not all. generally 70 to 80 percent. 2010 or Jiang et al. All of the above approaches are examples of “per-pixel” classification.e. Image segmentation. or the analyst while developing the training data sites. of the variables to build the resulting tree and identifies resulting splits based only on this subset of variables.. Harper et al. could be particularly useful in areas where the spectral separation of vegetation types is limited. Other approaches mine all available data. Each of the above approaches can be applied to a single image at a time. is used to inform the classification. Hansen et al. a five-by-five pixel window. Segmentation generally represents an intermediate step prior to classification. 2008. 2011). Most recently. or to mosaics of images of the same area and time period. 2005.. such as Random Forests 27.. Friedl et al. randomly selects some. This enables a direct estimation of change and persistence from the multi-temporal imagery. providing additional information that can be utilized by the classification algorithm. Some recent semi-automated approaches use much more of the data archive than a single image from a start date and another one from an end date. for example.. In this approach. and generate many multi-temporal metrics. Replicability and analyst interaction versus automation In the case of estimating deforestation. which may be critical in cloudy areas. accuracies for land-use classes such as agriculture and grassland tend to be lower.berkeley. 27 http://www. Rodriguez-Galiano et al. or some other metric.g.0 – REMOTE SENSING OF LAND COVER CHANGE 104 .edu/~breiman/RandomForests/cc_home. i. many studies with analyst interaction have produced accurate estimates of national forest cover. such as “linear trend in red reflectance” or “maximum middle-infrared reflectance recorded since the initial date” (e. 2012).. Evans et al. They can also be applied to multi-temporal image data. In contextual classification approaches..g. rather than the classification of two individual images and two single-date classification outputs. a pixel is classified based on its own spectral characteristics as well as those of surrounding pixels.. Another type could use the average. This approach. e. the variance of the pixels within a certain window around the center pixel.maximum likelihood and other classification algorithms. such as the entire Landsat archive. Longépé et al.. 2007. another contextual approach. This process also includes a single classification step that yields a two-date classification. and generally perform efficiently. These are powerful because short-lived signals of land-use change are more likely to be captured. unlike DTs. These errors would be compounded when the two maps are combined during post-processing. both of which may contain errors. Such methods do not suffer from overfitting. the central pixel being classified. Gislason et al.. 2012). 2008b). One type of contextual classification is textural classification. Random Forest ensemble classification methods have been successfully applied to land-cover and land-cover change classification (Pal. of the pixels within the window.g. and all available data are employed. Accuracies have often been reported over 90 percent (e. and segmentation algorithms allow an analyst to specify the relative size and shape of the segments. Weighting can also be used to apply different weights to pixels that are closer to.. and these estimates are generally derived from local rather than national studies. The resulting segmented image can then be classified at the segment level.stat. Any of these methods can be expanded to become part of a contextual classification. 2006. is a statistical method that groups contiguous pixels into areas (segments) that are relatively homogeneous.htm REDD+ MRV MANUAL: CHAPTER 5.. Some form of “direct change estimation” process is usually recommended for change estimation. images from the beginning and end of a study period. These approaches may be based on the seasonal signal of different types of vegetation and estimate changes based on where anomalies in these seasonal signals are detected (see. for example.

note that the previous section recommends directly estimating change from multi-temporal images. clear-cutting of forest) while applying a less-automated method to estimate the less conspicuous ones. Another approach could be to automate the estimation of the most conspicuous changes (e. Another example of automating the classification step could be to automate the process of collecting data on training sites. While the results are encouraging. 2008a). such as dark or clear water. and the cost. In doing so. A large set of training sites could be built for the entire country. This is often not the case and should therefore be tested. Countries could potentially automate all but the final steps of a methodology to estimate change. like areas with modest topography and cloud cover. Masek et al.In recent years there has been valuable research on automated methods for processing satellite data.. this indicates there is no single best answer and countries should evaluate options themselves. a country could choose to monitor certain parts of the country that are more appropriate for automated monitoring..0 – REMOTE SENSING OF LAND COVER CHANGE 105 . the following step should be applied to the two classifications to create a two-date change map. this should be done first. for example. and bright non-forest areas such as urban areas and exposed sand and soil. While rapid estimation of cover and change based on these rules could be achieved for much of the country. classification algorithms could be rule-based. or to classes that are spectrally distinct.. Once it is confirmed that this set can be used to produce an accurate map. in the form of thresholds applied to reflectance data. Further. especially mountainous regions or areas with more deciduous vegetation. With REDD+ MRV MANUAL: CHAPTER 5. further testing should be conducted in other regions. and merging can be accomplished by recoding the values of all the sub-classes to a value that represents the final class. then the actual change estimation is conducted using a set of rules or digital classification assisted by analyst interpretation (Souza et al. must be valid. Difficult areas may require more direct analyst interaction to obtain accurate results.g. there are well-published approaches that use automation for a series of pre-processing steps. snow. the recently published Deforestation Atlas of the Democratic Republic of the Congo was produced by an entirely automated approach (Hansen et al. (2008b) sampled an existing vegetation map to generate training points. the assumption that accurate results can be achieved over large areas. Conversely. Such rules. the same training sites could be applied to new data in later years to calculate new class spectral signatures to be used with the new imagery. A related approach is to use traditional interpretation methods to identify training sites for classes. 2005. may only yield accurate results over certain parts of the study area where the cover types are most easily distinguished with the spectral data. the validity of the results would be very dependent on precise correction and normalization of the images in the pre-processing step. and the post-processing steps required will vary depending on the classification approach and attributes desired in the final map product used to calculate areas for categories. The values of the classifications from both dates should be recoded to form the basis for a final class map that records categories of change and persistence. countries should seek an optimal balance between the accuracy of the final estimates. the replicability of the methodology. For example. In this case. if relatively few rules are used. There are a wide range of options to apply automation in the classification step itself. For example. The approach could be automated once a national training data set is defined. As the scientific community itself uses a broad range of approaches. Alternatively. If a two-date classification was not conducted. If the classification methodology included the creation of sub-classes merged into the final desired classes. although in some cases it has also included the classification step. This has mostly been in the pre-processing steps.. Post-processing Post-processing refers to any step conducted after the classification step. or for various strata within it. the remaining areas or classes would need to be estimated via other approaches. as the spectral variations in the new data are accounted for each time these new data are combined with the training site locations. Hansen et al. Each class in the digital output file of the classification has an assigned number. 2008). However. as is typically done in a supervised classification approach. using few rules. or derived data in some other units.

If reporting in absolute units. but they are common artifacts that should be removed. second. if the aim is to report in units of percent per year. many studies that report changes over five or ten years use images that are within one or two years of each target date. which may or may not be desired. The second type of filter is a “sieve filter. and can be used in sequence. this is useful because the final product can have a defined MMU that meets a country’s national definition of forest. These areas should be defined and recorded as having an eight and twelve- year difference in dates. First. If reporting in units of percent. patches of cells with the same value are identified. which may be the great majority of the study area. the output classifications will already have values representing classes of change and persistence and. this is subtle.” are based on the class values around (within a three-by-three window) a center cell. one should consider if the sampled area. in years (Puyravaud. After recoding to the final class values. and patches smaller than a user-defined size eliminated. not in per-year units. to eliminate patches smaller than a desired MMU or the minimum patch size in a national class definition. This allows for a temporal extrapolation for each area in each stratum. This may warrant another stratification so that percent rates are not extrapolated into very different areas and thus not well-represented by those where data exist. the total rate should be reported for the entire time period. For small windows. This is calculated based on areas at the start and end date: Annual change rate = [ (area t2 / area t1) ˄ (1 / (date t2 – date t1)) ] – 1 where t1 and t2 are the beginning and end of the time period of the study. In this case. Data change detection. the source images may not be from the exact beginning and end of the time period reported. each forest stratum experiencing change between the eight year difference would have an entry. especially where cloud cover limits the coverage of optical images. errors present in each of the single-date classifications would be compounded in the merged classification output. A common type is a local majority filter. yielding a two-date classification. First. REDD+ MRV MANUAL: CHAPTER 5. Further. As mentioned above. For example.0 – REMOTE SENSING OF LAND COVER CHANGE 106 . in this case to a ten year period. is representative of the entire study area. then a correction must be applied. therefore. This not only removes the speckled pattern but also smoothes jagged edges. For example. to eliminate small errors associated with micro-topography and other very local effects that produce a speckled pattern of mis-classified raster cells and. from cloud cover or other reasons. the study areas should be divided into areas where the image pairs representing the start and end time have different lengths of time separating them. the previous recoding step would not be required. Note that we use the term cell instead of pixel when referring to classification outputs rather than spectral images. one part of the analysis may be based on images from 2001 and 2009. After extrapolating the percent rates. and rates of change calculated for each area with a given difference. Calculating change rates Several factors must be carefully addressed when calculating change rates. for a 2000 – 2010 study. “Local filters. Some classification methods are more prone to producing these errors than others. rates can be converted to absolute values by combining with the forest area at the beginning of the reporting period. while another is from 1999 and 2011. such as a three-by-three window. Filtering is generally performed for two reasons. where the center cell is re-assigned to the most common cell value within the window. extrapolation is needed. and for each forest stratum. data on change can be combined with those on the forest strata. and each forest stratum experiencing change between the twelve year difference would have an entry.” In this. In the above example of 2000 – 2010. Two broad types of filters are commonly used. in order to report change for each forest stratum. 2003). some filtering of the product is usually desirable. if the analysis was not for a single-year period. must also be addressed. In this case. In this step. Third.

000 = 94.040 4.890 / 2. and associated statistics (Congalton. The statistics include overall accuracy.6 Forest 94. This is based on errors of commission. In the example error matrix in Table 5. Performing an accuracy assessment of a thematic map represents a very important component of the process.5 Non- Non-forest 98.000 Producer's accuracy (%) User's accuracy (%) Forest 99.930 4. This statistic is based on errors of omission. 2011. and thus accuracy assessment also refers to assessment of error or uncertainty. i.1 REDD+ MRV MANUAL: CHAPTER 5.6 percent.3: Example of an error matrix and resulting overall accuracy.8 Degraded Degraded Forest 92. 2012. Reference Forest Degraded Forest Non-forest Map total Forest 9..960 2.040 = 92.000 Land-use map Degraded Forest 70 1. An accuracy assessment allows errors in the map to be estimated and uncertainty quantified. In this hypothetical case. thus providing additional explanation of.880 90 30 10. and validity to. Overall accuracy is the portion of the total number of correctly mapped pixels. The user’s accuracy indicates how often a pixel was incorrectly assigned to a given class. The error matrix is generated by comparing the classification results with reference data. . how often a pixel was incorrectly included in a class. The main elements of an accuracy assessment are the error matrix.. The user’s accuracy for the same class is: 100 x 1.000 Reference total 9.000 16.e. Foody.3 forest 98. or confusion matrix. The producer’s accuracy indicates how often a pixel is correctly assigned to a specific class.e. and the producer’s and user’s accuracy for each class in the product.5 percent. The Kappa coefficient can also be calculated. the columns contain verified land uses and the rows contain estimated uses from the classification. In the example table.890 40 2. Pontius and Millones. The values along the diagonal are the number of correctly classified pixels. i.2 Forest 98.3 Table 5.0 – REMOTE SENSING OF LAND COVER CHANGE 107 . in this case the estimate of an area of change over time. but many articles highlight the limitations of this statistic (Olofsson et al. the results. the land use totals are the same as in the beginning time in Table 5. Accuracy equals one minus the error value.3.890 / 2. 1991). and those off the diagonals are errors of omission and commission.. the producer’s accuracy for degraded forest is: 100 x 1.000 Non-forest 10 60 3. 2002). how often a pixel was incorrectly omitted from the class. Accuracy assessment Uncertainty is the error in a particular estimate.

It also allows one to assess where improvements are most needed. Once the error for a particular area of change. This type of proportional sampling design ensures that adequate sampling occurs in sparser yet critical classes. and iii) analysis. which answers the question.3 demonstrate error estimation in land-use cover for a single date. the data can easily be transferred. ‘what is a suitable subset area to sample?’. This is especially important as it quantifies the confidence of a particular class. While they differ from the reporting-table format of the IPCC. The examples in Table 5. The sampling design could focus on areas of deforestation identified in the map. REDD+ MRV MANUAL: CHAPTER 5. This can be done using Equation 5. Multiple interpreters can be used.e. which parts of the GHG inventory are contributing the greatest errors in GHG estimates and should be reviewed as part of the KCA. countries must estimate errors in land-use change estimates over time. cross-checked. and a stratification could be used to categorize areas of high change probability and low change probability. 2010). It is useful to calculate these statistics for the different strata in a study area. such as deforestation. and the consistency of their interpretation can indicate confidence of the validation data set itself. Section 3. (2013) describe a process for creating CI bounds based on area-adjusted error matrices In addition.. Manual interpretation of an image by an analyst is generally considered higher quality than an automatic classification algorithm). provides guidance on considerations for generating reference data and performing an accuracy assessment.e.2 from the UNFCC “Good Practice Guidance for LULUCF” (2003) for the propogation of errors. is estimated. visual interpretation of a combination of very high resolution imagery. to generate area-adjusted errors based on the proportional area of each class and errors identified in the matrix. 2013). This will lead to stronger estimation of the overall uncertainties in estimated GHGs. similarly to the application described in Chapter 4. Accuracy assessment is very important. such as degraded forest.0 – REMOTE SENSING OF LAND COVER CHANGE 108 . shown at the end of Chapter 3. The GFOI MGD also provides two example approaches for performing an accuracy assessment and area estimation. which answers the question of ‘how to calculate accuracy and quantify uncertainty?’.. Field surveys can be valuable for the classes that are the most difficult to interpret even with very high resolution imagery. (2014) also provide a full review of good practice recommendations for producing transparent and “scientifically rigorous” accuracy estimates and estimates of area based on change between time one and time two. Olofsson et al.2. and ii) Combining reference data sources (i. However. Error-adjusted area estimates. together with the total area of each class identified in the map. which answers the question ‘are the maps and reference data in agreement?’. An appropriate approach is the use of careful. Proportional sampling could then be focused in these strata to ensure each class is adequately represented in the validation analysis. one stratified and one model-based approach. Olofsson et al. Error matrices and accuracy assessments can also be extended to provide confidence interval (CI) information. Multi-date accuracy assessments use the information available from two dates. it is important to account for a rare class when developing a validation strategy (Stehman et al. it can be combined with the error of the change in stocks per unit area estimated via field inventories. These include: i) ensuring the reference data are of a higher quality than the map data (for example. such as those described in Olofsson et al. field and aerial surveys which can be particularly cost effective if resources are limited). i. thus providing very pertinent additional information. and MRV programs should work with expert statisticians in developing strategies for validation sampling and combining information on uncertainties in AD with those in EFs. (2013). along with imagery used in the classification itself.. understanding that evaluation of the entire map is not possible. They detail three separate steps that should be undertaken to complete an accuracy assessment including: i) the sampling design.Both the land-use change and error matrices are common formats for reporting land-use change and errors. use the information available in the matrix. ii) the response design. as this allows one to combine errors with errors in carbon stock for each stratum. or AD.7 of the GFOI MGD (GFOI.

forest fires (canopy and sub-canopy) and fuelwood collection.5 EMERGING AREAS OF RESEARCH Several areas of particularly active research in support of REDD+ activities include: the mapping and monitoring of degradation.11: Examples of Landsat and RADAR images. Multiple definitions of forest degradation exist. Hirata et al. and the increased use of field sampling to facilitate remote sensing product validation. Mapping and Monitoring of Degradation Forest degradation is a substantial contributor to GHG emissions from land-use change (Nepstad et al. Mapping and monitoring of forest degradation remains challenging. A) B) Figure 5. 2005). Ground observations in the Landsat data in A) are partly obscured by clouds. Several characteristics make RADAR an attractive source of information for such applications. such as RADAR.. 1999. from Peru. GOFC-GOLD (2012) presents a range of human activities that result in forest degradation including selective logging. The IPCC’s definition of forest degradation is provided in section 5. 2012. Houghton and Hacker. deforestation. optical sensors. and a country should seek to understand the implications and applicability of different approaches. this is more noticeable in the latter REDD+ MRV MANUAL: CHAPTER 5. 2003.. they are able to penetrate clouds and are thus useful for monitoring in areas with persistent cloud cover. for example. 2012). Peru. Conversely. Lambin et al. Asner et al. because the signals received by the sensor are less affected by atmospheric conditions. adding to the complexity of mapping and monitoring forest degradation. forest fires (canopy and sub-canopy) and fuelwood collection.0 – REMOTE SENSING OF LAND COVER CHANGE 109 . in monitoring. and degradation.11 illustrates a detail of a Landsat image compared to a PALSAR satellite image for an area in San Martín. Stickler et al.2. RADAR images are directly comparable over time. for example.5. Readers should also refer to the relevant sections of both the GOFC-GOLD SourceBook and the GFOI MGD. 2005.. While the brightness variations in both are affected by terrain. GOFC-GOLD (2012) lists a range of human activities that result in forest degradation including selective logging. because RADAR sensors operate in longer wavelengths (generally 1cm to 1m) of the EM spectrum than.1. GOFC-GOLD. with estimates ranging from 20 to 50 percent of total land-use GHG emissions over large regions (see.. and the properties of the emitted radiation from active sensors are controlled and well known. RADAR signals are also sensitive to the geometric properties of a forest. providing information on the distribution of aboveground biomass. to map and monitor forest extent and characteristics. The use of other sources of remotely-sensed data in mapping and monitoring forest A second area of research in support of REDD+ activities is the use of other sources of remotely-sensed data. 2009. Figure 5. These different activities may require different monitoring approaches. such as RADAR. In addition. the use of other sources of remotely-sensed data. Souza and Roberts. 1999. First. while the PALSAR image in B) is cloud free.

A key characteristic of RADAR data not found in most optical data is polarization. that illuminate a strip of the earth’s surface (swath). REDD+ MRV MANUAL: CHAPTER 5. Apart from polarization. and • VH: vertical transmission and horizontal reception. side-looking observation. Water bodies tend to be relatively smooth. several additional characteristics distinguish RADAR instruments and data from their optical counterparts and are useful to understand. geometric shape and dielectric properties of an object also affect the information received by the RADAR sensor. In active remote sensing.5 Polarization refers to the orientation of the electric field with respect to the direction of propagation. whereas a P-band (λ= 23 cm) may penetrate leaves and small branches thus providing information about both the big branches and stems of the trees.The following provides a brief introduction to a selection of active remote sensing concepts. this is described below. and these are strongly affected by their moisture. For example. the surface with a higher moisture content will appear brighter. For example. For soil types of similar roughness. Linear is the most common polarization used in RADAR remote sensing where a radiated electric field is oriented either horizontally (horizontal polarization) or vertically (vertical polarization) with respect to the direction of propagation. The below discussion includes a selection of SAR applications that are relevant to REDD+ activities. such as an X-band (λ= 3 cm) may only penetrate the upper layer of a forest canopy. as shown in Figure 5. As with optical sensors. A shorter wavelength band. These include phase. including RADARSAT-2.12. causing backscatter. but smooth for longer wavelength RADAR such as P-band RADAR. the electric field of the resulting radiation has a preferred orientation. • HV: horizontal transmission and vertical reception. small objects such as leaves and twigs are considered rough for small wavelength RADAR. while other satellites. Thus. and thus have a bright appearance in a RADAR image. • VV: vertical transmission and vertical reception. RADAR sensors exploit different wavelength bands. ENVISAT and ALOS/PALSAR acquire data with all four polarizations (“quad-pol”) or two polarizations (“dual-pol”). with most of the energy being reflected away from the RADAR. Synthetic Aperture RADAR (SAR).0 – REMOTE SENSING OF LAND COVER CHANGE 110 . while trees and other vegetation are rough. and polarimetry. and the distance is calculated based on the time elapsed for the RADAR signal to make a complete round trip. Surface roughness is a relative term that is dependent on the RADAR wavelength. RADAR measures the distance between an object on the ground and the sensor based on the strength of radio waves that are transmitted as pulses of microwave beams. resulting P-band images are important for measuring vegetation biomass and aboveground carbon stocks. The intensity of the signal that is scattered back to the receiver from this transmitted energy is recorded as the returned signal. The surface roughness. The difference in intensity of RADAR returns from two surfaces of equal roughness is an indication of the difference in their dielectric properties. The next transmitted pulse illuminates the next strip of terrain along the swath. and a two-dimensional image is created (each pulse defines one line). Additional Such concepts are key to understanding the basic characteristics of active remote sensing data. Some space-borne satellites including RADARSAT-1 and ERS-1/2 have only single polarization (RADARSAT-1 with HH and ERS-1/2 with VV). the brightness of areas covered by bare soil may vary depending on the roughness and moisture content of the soil. A sensor that can transmit either horizontally (H) or vertically (V) polarized waves and receive both will result in the following four polarized images: • HH: horizontal transmission and horizontal reception. interferometry. these are outlined in Box 5. directed by an antenna.

such as Synthetic Aperture RADAR (SAR) systems. a RADAR image contains information about the intensity of the signal and the phase. or lag.e. If two SAR images have been acquired over the same area from very close antenna positions. ortho- rectified. this grainy ‘salt and pepper’ texture degrades the quality of the image and complicates interpretation. while some geometric and radiometric distortions due to terrain relief may persist. therefore.12: Horizontal and vertical polarizations Box 5. these processed and derived products are generally more appropriate for use in mapping. Speckle can be reduced by averaging the backscatter response within a pixel. Additional characteristics of RADAR Phase describes the relationship of the lead. observing a location directly below the sensor). As previously discussed. the observed phase differences can be used to infer three- dimensional information about the terrain height. 360 degrees represents one complete cycle and. containing both amplitude and phase information. Since the antenna positions are precisely known. a signal that reflects off a tree trunk to the ground surface is likely to show distinctive polarization shifts from signals that return directly off the soil. This across-track capability. and radiometrically corrected. layover. termed Side Looking RADAR (SLR). The path difference is geometrically related to the distance between two antennas and the terrain height. more parameters can be measured from polarimetric RADAR compared to single-channel RADAR. a wave that is lagging one quarter of a wavelength behind the reference has a phase of 90 degrees. Among the distortions that may persist is speckle. introduces a range of geometric distortions including foreshortening. Another relevant RADAR concept is SAR interferometry. SAR speckle causes pixel-to-pixel variation in intensities even over a homogenous area. The technique is known as SAR Interferometry. SAR polarimetry is a relevant RADAR concept. For example. REDD+ MRV MANUAL: CHAPTER 5. Finally. and is expressed in degrees.0 – REMOTE SENSING OF LAND COVER CHANGE 111 . Such unique information is important for discriminating different land-cover types. or by applying smoothing filters. the different path lengths from these positions to the object on the earth’s surface cause the differences in phase. Some RADAR remote sensing systems. though this can effectively reduce resolution.5. And. correction. As mentioned in the introduction. unlike many optical sensors which acquire imagery at nadir (i. Data from such systems are generally processed by data distributors to Single Look Complex level data. Most RADAR sensors are also side-looking instruments. The different polarization bands may contain unique and additional information about the surface object.. are able to achieve a relatively high resolution without the use of a large antenna. of an electromagnetic wave with respect to a reference wave of the same wavelength.5. Surface objects that scatter are vertically oriented and show high backscatter in vertically polarized imagery and low backscatter in horizontally polarized imagery. or partial. and a range of derived products which are usually geocoded. and shadows that require full.

One major limitation of SAR data utilization and analysis is the difficulty involved in interpreting RADAR backscatter as compared to optical spectral data (Saatchi et al. The recent systematic availability of fully polarimetric SAR data from the ALOS-PALSAR. and complex REDD+ MRV MANUAL: CHAPTER 5. 2009). ENVISAT. the first long wavelength (L-band. with an emphasis on characterizing several clearing practices and forest regeneration characteristics.. 2000). They also mapped forest patches and fragmentation and found these data helpful in delineating areas with different degrees of forest disturbance. Landsat. (2012) generated biomass maps and changes in carbon stocks with known uncertainties using PALSAR imagery in a region in central Mozambique yielding maps with sufficient accuracy to enable changes in forest carbon stocks of as little as 12 tons per hectare over 3 years. 2000) and approximately 81 percent of global forests are above this saturation limit (Nelson et al. producing overall accuracies of 92 and 94 percent with PALSAR and Landsat. Brazil. SAR data are also being evaluated for scaling up ground-based AGB and monitoring changes over large scales (Lu. 2006. for the forest versus non-forest classifications. Mitchard et al. They also found a high degree of spatial similarity among maps derived from PALSAR. (1997) used C-SIR data to map land-cover types and monitor deforestation in the tropics. several limitations exist and the history of SAR data usage for land-cover classification remains relatively recent. Saatchi et al. while the Landsat TM yielded a more accurate deforestation extent classification. characterized by small scale deforestation and degradation. Substantial improvements in land-cover change classification can be achieved by combining polarimetric and polarimetric interferometric information (Shimoni et al. (2011) used L-band synthetic aperture RADAR backscatter data from 1996 and JERS-1 and PALSAR data from 2007 to produce biomass maps of a forest–savanna ecotone region in central Cameroon. this sensitivity appears to saturate at biomass levels of around 100 tons ha-1 (Imhoff et al.Applications of SAR Applications on the use of SAR data for forest mapping. The presence of topographic effect and speckle complicates both visual and digital analysis of RADAR images. For example. 25 cm) SAR satellite with the capability to collect cross-polarized responses. 2007). and RADARSAT-2 has led to further land-cover classification research using SAR imagery. has yielded improved estimates of AGB with little or no saturation.. However.. 2009). (2012) used a combination of PALSAR. 2006).. Ryan et al. While these results highlight the potential for space-borne imaging RADAR for estimating forest area and biomass.0 – REMOTE SENSING OF LAND COVER CHANGE 112 .. In addition to polarimetric information. (1997) compared SIR-C data with Landsat TM in a test site in Rondonia. Mitchard et al. They found that the RADAR data detected changes in a broad AGB class in forest–savanna transition areas with an accuracy of 95 percent. the fusion of spatial and textural information derived from various SAR polarizations has been shown to improve classification results (Borghys et al. such as trees. and ground-based data to map AGB in Gabon’s Lopé National Park. and existing data from Projeto De Estimativa De Desflorestamento da Amazonia (PRODES). respectively. (2010) assessed the ability of PALSAR and LANDSAT data to classify and map forest cover in the Xingu River headwaters in southeastern Amazonia. Walker et al. Recently. In addition. Rignot et al. Multiple studies have tested the potential of combined RADAR channels of different frequencies and polarization for deforestation monitoring.. Mitchard et al. space-borne LiDAR (ICESAT GLAS). measuring and monitoring aboveground biomass (AGB) and scaling-up ground-based AGB measurements are increasing. with 95 percent confidence to be detected. the combined use of both Landsat and RADAR imagery further improved the mapping accuracy. the Brazilian Amazon deforestation monitoring program. up to 250-300 tons per hectare based on the sensor’s cross-polarized ability to exploit the strong response of three dimensional objects. polarimetric interferometric SAR (PolInSAR) provides interferometric information (see Box 5. compared to bare soil. These data are sensitive to the geometric properties of the forest and directly related to measurements of AGB. data from PALSAR. Similarly.5) related to the structure and complexity of the observed objects.

Field measurements should also be collected as close as possible to the acquisition date of the remote sensing image. reducing the impact of GPS inaccuracy (Mascaro et al. the Japanese Space Agency (JAXA) launched ALOS-PALSAR-2 in May 2014. Additional evaluations are also needed to assess the utility of newer RADAR data sources in mountainous areas. This approach to assess model uncertainty is recommended because it is straightforward and captures the overall result of many potential sources of error that would otherwise have to be independently quantified and then propagated. This is of particular interest in the often vast and difficult-to-access landscapes contemplated under most national and jurisdictional REDD+ MRV. and classification methods based on polarimetric decomposition are being developed. 2013). used to assess the accuracy of the predicted values at different scales. Therefore. are no longer collecting. a sample of plots are measured and held in reserve. in particular high resolution satellite imagery. Note that these preferred sample designs require that the remote sensing data already be in hand. for example. and any RADAR or LiDAR-based estimates of biomass will need to be calibrated using field-plot data. RADAR. Circular plot designs and a minimum plot size of 0. This can be achieved via weighted random sampling. these data can be valuable in extrapolating field-based estimates over larger areas. and various kinds of LiDAR hold promise to more efficiently map biomass stocks. other unequal probability sampling approaches or a systematic sample approach. as the objective is not to produce a field- based estimate but instead to model predictions. These data likely will not provide more accurate biomass data for plot sites than those derived from field data. The sample should achieve a roughly even distribution across the range of conditions of the remotely-sensed data. they are still in the process of refinement. Uncertainty related to the long-term data continuity of space-borne RADAR systems could also prove a limiting factor for forest monitoring. REDD+ MRV MANUAL: CHAPTER 5.2 ha is recommended for use with airborne LiDAR to reduce model errors (Zolkos et al. Sampling strategies differ however. Although. and after model construction and application. especially remote areas that are costly to access. In the latter case. though the range of advanced SAR processing techniques capitalizing on the availability of multi- polarimetric information is evolving.areas with a greater abundance of secondary forests may yield significantly lower accuracies.0 – REMOTE SENSING OF LAND COVER CHANGE 113 . While these approaches hold promise. such as the range of reflectance values. providing an equivalent range of spectral values within the area sampled via remote sensing. from remotely-sensed data. Large. Field measurement plots can calibrate and validate prediction models. Use of Field Sampling To Facilitate Remote Sensing Product Validation Remote sensing data. 2011). Larger plots facilitate the alignment of ground samples and remote sensing imagery. Finally. However. and this instrument again contains an L-band. which both provided fully polarimetric L-band data. to optimize the design... SAR data generally yield less robust results than Landsat data for forest/non-forest classification in most studies. over a wider spatial extent. where significant relationships can be calibrated. fixed-area plots are best suited for generating ground-reference data to compare with remotely-sensed data. The same field measurement and QA/QC procedures described in Chapter 4 are also relevant in the collection of ground truth data. PALSAR and ENVISAT. ideally the sample should be well-distributed across the area to which the model will be applied. However. ground-reference data used to build prediction models for remotely-sensed data do not need to be strictly uniformly sampled.

T. A method for integrating MODIS and Landsat data for systematic monitoring of forest cover and change in the Congo Basin. Laporte. Huang.F. G. Adusei. Hackler.M. A Review of Assessing the Accuracy of Classifications of Remotely Sensed Data. Justice and A. gains and losses of carbon stocks in forests remaining forests. A sourcebook of methods and procedures for monitoring and reporting anthropogenic greenhouse gas emissions and removals associated with deforestation. Hansen M. Schneider. Switzerland. Supervised feature-based classification of multi-channel SAR images. 2006.J. Y. 1999. http://www. B.G. MODIS Collection 5 global land cover: Algorithm refinements and characterization of new datasets. A. Transaction on Geosciences and Earth Observation 37(2): 969-977. 1991. Yvinec.R. and X. Random Forests for land cover classification. Roy. Walker. J. A. Roberts. Pizurica and W. W. Costa. 2005. Perneela.C. 2006. Maximizing Land Cover Classification Accuracies Produced by Decision Trees at Continental to Global Scales. Telmer and T.. Brodley. N. A. FAO. Clark. and forestation. Broadbent. 2010. E. GOFC-GOLD Report version COP19-2 (GOFC-GOLD Land Cover Project Office. Remote Sensing Environment 80: 185- 201. Choosing a Forest Definition for the Clean Development Mechanism. J. K. A.. C.C. R.. Samanta and R. Friedl.fao. Forests and Climate Change Working Paper 4. Remote Sensing Environment 112: 2495–251. D. Borghys. P. C. R.R. and A.S.A. D. 2008a. Remote Sensing of Environment 91(1): 68-89. M. Silva. Sveinsson. Small-footprint lidar estimation of sub-canopy elevation and tree height in a tropical rain forest landscape. Philips.L. Houghton.A. Ramankutty.. Foody.. P.0 – REMOTE SENSING OF LAND COVER CHANGE 114 . Sun. Silva.N. G. Altstatt. Selective logging in the Brazilian Amazon. M. Sulla-Menashe. Oliveira.A. Status of land-cover classification accuracy assessment.5. N. S.P. Dubayah. Knapp. 2002.php. D. Wageningen University. M. Global land cover classification at 1 km spatial resolution using a classification tree approach. REDD+ MRV MANUAL: CHAPTER 5.E. D.S DeFries. Available at http://www. 2010. 2012. Remote Sensing of Environment 37: 35-46. 2006.. Goetz. J. Group on Earth Observations. GOFC-GOLD.J. D. Lindquist.G.O. P. 2000. M. The Netherlands).O. 2013. Benediktsson and J. Hansen M.L. Science 310: 480–482.N. C. Strahler. J. Friedl. Baccini. Clark. Gislason. Integrating remote-sensing and ground-based observations for estimation of emissions and removals of greenhouse gases in forests: Methods and Guidance from the Global Forest Observations Initiative. 2004.C. Remote Sensing of Environment 114: 168-182..wur.B. M. A.gofcgold. Geneva. S.. T. 2013. Congalton. Pattern Recognition Letters 27: 294-300. International Journal of Remote Sensing 21(6): 1331-1364. R.6 REFERENCES Asner. Beck. M. A.. Sulla-Menashe. and D. Tan. Pattern Recognition Letters 27: 252–258. Sohlberg.E.H. Friedl. Estimated carbon dioxide emissions from tropical deforestation improved by carbon-density maps.P. B. Using ALOS/PALSAR and RADARSAT-2 to Map Land-cover and Seasonal Inundation in the Brazilian Pantanal. Keller and J. Evans. IEEE Journal of Selected Topics in applied Earth Observation and Remote Sensing 3: 560-570. Nature Climate Change 2: 182-185. D. Townshend and R. Sibley. S. A. GFOI. M.

2012. Remote Sensing Environment 115: 3770–3774. Roy and C. Huang.V. 2003. Forestry and Forest Products Research Institute Japan.P. C. Pipatti. K. Potapov.). Ngara. Gytarsky. G. Flood.J.A.C. D.. Isoguchi. Cohen. Tanabe. Y. 2007. Stolle.J. H. Detto. Takao. Lawrence. North American Forest Disturbance Mapped from a Decadal Landsat Record. Toriyama. Hansen. Asner. Zhu. A Simple Semi-Automatic Approach for Land Cover Classification from Multispectral Remote Sensing Imagery. REDD Research and Development Center. M. Treitz. Wulder. K. P. M. IEEE Transaction on Geosciences and Earth Observation 49: 2135-2050. D. Hirata Y. P. Geist. J.J. Progress In Physical Geography 27(1): 88-106. K.L. Sato. (2008). Imhoff. and Nelson.G. Steininger. F. S. M. M. Remote Sensing Environment 112: 2914–2926. Carroll.K. Townshend. L.. IPCC. and F.K. International Journal of Remote Sensing 29: 7269-7275. Rakwatin.R. Tucker.S. Masek. Zhuang.. E. BioScience 52(1) 19-30. Kanagawa (eds. Johnson. Yulianto. Published: IGES.0045889 Lambin. M. Mascaro. J. WB. Chicester.S. K. PLoS ONE 7(9): e45889. P. B.C. P.R. Holford. W.F. doi:10. W. T... Lefsky.P. 156pp. Harper. Shimada. P. Miwa. Mather. R. IPCC GPG-LULUCF. Fifty years of deforestation and forest fragmentation in Madagascar. The potential and challenge of remote sensing-based biomass estimation. 2003. G. Huang.C. Transaction on Geosciences and Earth Observation 38: 1458–1463.0 – REMOTE SENSING OF LAND COVER CHANGE 115 . IPCC. 2000.. Published: IGES Japan. Buendia. J. Jiang. Tanabe and F. Computer Processing of Remotely-Sensed Images. D. Muller-Landau. Y. Loveland.. BioSAR™: an inexpensive airborne VHF multiband SAR system for vegetation biomass measurement. L. M. M. Wolfe. D. M. International Journal of Remote Sensing 27: 1297-1328. The suitability of decadal image data sets for mapping tropical forest cover change in the Democratic Republic of Congo: implications for the global land survey. Ngara and K. 2011. G. ISBN 978-4-905304-15-9. Parker and D. Humid tropical forest clearing from 2000 to 2005 quantified using multi-temporal and multi-resolution remotely sensed data. J..1371/journal. 2011. J.pone. Annual Review Environment Resource 28: 205–41. Stehman.. Longépé. eds.W.J. W. In: Penman. England. Lim. and C. Pittman. 2008b.G. And H. Ren. Lepers.. 2012. P. Good Practice Guidance for Land Use. L. eds.. Lidar Remote Sensing for Ecosystem Studies. Evaluating uncertainty in mapping forest carbon with airborne LiDAR. M.. Y. IPCC-IGES. F. J. X. Environmental Conservation 34: 325-333. Miwa. H. 2006 IPCC Guidelines for National Greenhouse Gas Inventories. Buendia K. Hyer. J.Hansen. T. REDD+ MRV MANUAL: CHAPTER 5. Harding. T. Wiley & Sons. G.V. M. Land-Use Change and Forestry. DeFries.O. Justice.. Cohen. Eggleston. Dynamics of land-use and land-cover change in tropical regions. R. T. Hawkins. and K. Kruger. D.. Prepared by the National Greenhouse Gas Inventories Programme. Japan. C. St-Onge and M. 2008. Lindquist E. Dimiceli. Krug. and E. 2003. Harcombe. D. 2006. N. PNAS 105: 9439-9444.. REDD-plus Cookbook. M. Lu. 1999. 2006. Steininger. Assessment of ALOS PALSAR 50 m orthorectified FBD Data for Regional Land-cover Classification by Support Vector Machines. Xu and H.. Definitions and methodological options to inventory emissions from direct human-induced degradation of forests and devegetation of other vegetation types. May... O. 2003.J. Hall. R. 2002. LiDAR remote sensing of forest structure. T. Uryu. Kutler.J. G. Wagner. P.. S.

Lewis. B. Naesset. Pal. S. Meir.E. S. Remote Sensing Environment 129: 122-131. and P.Mitchard. T. E. R.A Lefsky. K. Nepstad. Hyde.. and W. Woodhouse. I. C. S. Forest Ecology and Management 177: 593-596. E. Canadian Journal of Forestry Research 39: 862-881. Skole.A. Investigating RADAR-LiDAR synergy in a North Carolina pine forest. Measuring biomass changes due to woody encroachment and deforestation/degradation in a forest- savanna boundary region of central Africa using multi-temporal L-band RADAR backscatter. Abernethy.M.S. E. I. E. Random forest classifier for remote sensing classification. Cochrane and V. A. Stehman. Saatchi. Lewis. Global Change Biology 18(1): 243-257.G. Cassells. Olofsson.. Lefebvre. Remote Sensing Environment 86: 554-565. H. Rignot. T.T.M.S.V. Mitchard. Feldpausch and P. 2011.. E. Feldpausch. Remote Sensing Environment 59: 167-179. 2009. P. Gobakken and G. Salas and D.E. Mitchard. Lewis. M. Mapping tropical forest biomass with radar and spaceborne LiDAR in Lopé National Park. Large-scale impoverishment of Amazonian forests by logging and fire. Ribeiro. B. S.. Ryan.. Miller. A. Ghee. An assessment of the effectiveness of decision tree methods for land cover classification. Mendoza. Woollen. E. E. G. 2003. Woodcock 2013. C. 2005. Leal. Woodhouse. Olofsson. Fischer 2003. E. M. C. Woodhouse and M. Verissimo. Schlesinger. Imhoff. P. Rodriguez-Galiano. P. S.M. Alencar. I.H. E.S. Meir. Mather.. Campbell.F.J. Congalton. and R. 1997. Remote Sensing Environment 110: 98–108 Nelson. S. Stehman and C. White. Rogan. 2003. Rowland. Chica-Olmo. M. Good practices for estimating area and assessing accuracy of land change. M. K. M. W. Saatchi. Potter. Collins.J.T. S. Nangendo. Quantifying small-scale deforestation and forest degradation in African woodlands using RADAR imagery. D. REDD+ MRV MANUAL: CHAPTER 5. Saatchi. Grace. G. M.. L. R. Puyravaud. Pal. Pardo-Iguzquiza. Emessiene. T. Standardizing the calculation of the annual rate of deforestation. Levian and C. Woodhouse P. G. P. N. A. Lima. Franklin. Sonk´e. Hill. Brooks. using imaging RADAR and thematic mapper data. M. Williams.L. C. Woodcock. T. Herold. Land-Cover Change Monitoring with Classification Trees Using Landsat TM and Ancillary Data. Photogrammetric Engineering & Remote Sensing 69(7): 793-804. T. and P. A. Nobre. D. P. M. P. Mitchard. Making better use of accuracy data in land change studies: Estimating accuracy and area and quantifying uncertainty using stratified estimate. J. I. Estimating Quebec provincial forest resources using ICESat/GLAS. Remote Sensing Environment 148: 45-57.H. Jeffery. V. Nature 398: 505-508.L. 1999. Photogrammetric Engineering & Remote Sensing 78(2): 129-137.. Moutinho. J. E.M. S. Geophysical Research Letters 36. Brazil. T. 2012. J..L. Wulder 2014. T. Incorporating the Downscaled Landsat TM Thermal Band in Land-cover Classification using Random Forest.. C. G. P. M. S.P. L. B. Stahl. Mapping deforestation and secondary growth in Rondonia. 2009. J. J.. S. T. Williams. Stow. C. Ryan. Remote Sensing Environment 115: 2861–2873. Ghimire.H. Biogeoscience 9: 179-191. M. Margolis. and Meir. Johnson.F. Foody. Foody. Edwards 2007.E.A. M.R. Boudreau. 2012.M.V. International Journal of Remote Sensing 26(1): 217-222. 2012. A. Gregoire. Nelson R.G. J. C. Using satellite RADAR backscatter to predict aboveground woody biomass: A consistent relationship across four different African landscapes.0 – REMOTE SENSING OF LAND COVER CHANGE 116 . Gabon: overcoming problems of high biomass and persistent cloud.

Buermann. J. Woodcock. Sibley.L. W. Alger. Soares and D. E. Benchmark map of forest carbon stocks in tropical regions across three continents. UN-REDD Programme. G. International Journal of Applied Earth observation and Geoinformation 11: 169-180. and J. Saatchi. M. 2005. S. 2010. Lenney and S. Delabie. Musinsky. S. Stickler.M. D. Macomber. J.V.. Harris. C. 2011. K. A meta-analysis of terrestrial aboveground biomass estimation using lidar remote sensing.S.0 – REMOTE SENSING OF LAND COVER CHANGE 117 .Saatchi. Mapping forest degradation in the Amazon region with Ikonos images. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing 3: 594-604. Roberts. Goetz and R. Reporting and Verification (M & MRV) in the context of REDD+ Activities. M. Hagen. L. H. B. Claudia. 2000. Woodcock. 2001. Accuracy 2010 Symposium. S. 2005. Coe. Kellndorfer. Olofsson. Morel. Alves. D. C. Remote Sensing Environment 59: 191–202. The potential ecological costs and cobenefits of REDD: a critical review and case study from the Amazon region.T. Borghys. K. and D. K. Lewis. Fusion of Pol SAR and PolInSAR data for land-cover classification. M. M. S. Shimoni. National Forest Monitoring Systems: Monitoring and Measurement. M..G. Stehman.. Silman and A. C. Republic of Congo. W. Acheroy.C.S. W. Moghaddam. N. Estimation of crown and stem water content and biomass of boreal forest using polarimetric SAR imagery. Brown. Leicester. Global Change Biology 15: 2803–2824. Combining spectral and spatial information to map canopy damages from selective logging and forest fires. Kirsch and D. 1997. Saatchi. 2013. D.S. Zolkos. S. D. Newell. Salas. 2009. Large-Area Classification and Mapping of Forest and Land-cover in the Brazilian Amazon: A Comparative Analysis of ALOS/PALSAR and Landsat Data Sources. and M. 2009.A. UK..L. PNAS 108(24): 9899-9904. Cochrane. A. Walker. Soares-Filho and E. July 20-23.R.. Lefsky.A. Mapping deforestation and land use in Amazon rainforest by using SIR-C imagery. Brazzaville. Nepstad. Classification and Change Detection Using Landsat TM Data: When and How to Correct Atmospheric Effects? Remote Sensing Environment 75: 230-244. Mitchard. IEEE Transaction on Geosciences and Earth Observation 38(2): 697-709.M.A. Friedl. 2001. Agosti. P.S. B.. J. Designing a reference validation database for accuracy assessment of land cover. R. Souza. S.C. Jr. White. Examining Fragmentation and Loss of Primary Forest in the Southern Bahian Atlantic Forest with RADAR Imagery. W. 26-27 October 2012. Dubayah..P. S. M. C. Song. Heremans. Zutta. S. Roberts and M.. Davidson. S. Walker.C.S. Petrova.J. J. C. D.T.A. Remote Sensing Environment 98: 329-343. Conservation Biology 15(4): 867- 875.. Souza. Herold 2010. C. Mcgrath. Sulla-Menashe and M. Saatchi. C. Perneel and M. M. D.O . Nepstad.. Seto.. S. M.M. Rodrigues. S. REDD+ MRV MANUAL: CHAPTER 5. International Journal of Remote Sensing 26(3): 425-429.E. Remote Sensing Environment 128: 289-298.

log ponds SPOT-6 Airbus Defense & 60km 6m multispectral Varies No Yes deforestation. 5m panchromatic (2. roads. roads. Space 1. 185km 15m panchromatic 16-21 days Yes yes deforestation. HRC CBERS-2B – INPE 27km 2.NASA 60km 15m multispectral Varies No Partial (no SWIR deforestation. NASA 30m multispectral encroachment. NASA 30m multispectral gaps encroachment.7m panchromatic 26 days Yes Yes skid trails. 165km 15m panchromatic 16-21 days Yes Yes. log ponds ASTER Terra . interpolated) roads. log ponds SPOT-5 CNES 60km 20m multispectral Varies No Yes deforestation. 60m thermal roads. canopy gaps. log ponds CCD CBERS-2B – INPE 113km 20m multispectral 26 days Yes Yes deforestation. encroachment. 1b Space multispectral logging roads REDD+ MRV MANUAL: CHAPTER 5.5m panchromatic encroachment. 100m thermal roads. illegal fishing vessels MODIS Terra / Aqua – 2330km 250m visible 4 times per Yes Yes fires.0 – REMOTE SENSING OF LAND COVER CHANGE 118 . log ponds Landsat 8 OLI LDCM . channels) encroachment.5m encroachment. GeoEye nadir 1m panchromatic (1) illegal fishing vessels / logging vehicles Pleiades 1a and Airbus Defense & 20km 50cm panchromatic 2m Varies No Yes Skid trails.7 COMMON SATELLITE DATA SOURCES FOR LAND-USE MONITORING Sensor Satellite – Swath Resolution Repeat Systematic Operational Monitoring Agency Width Cycle Acquisitions Status Applications Landsat 7 ETM+ Landsat 7 .3km @ 4m multispectral (4) Varies No Yes skid trails.5. with SLC-off deforestation. large-scale NASA 500m multispectral day deforestation 1km thermal (diurnal/ nocturnal) AWIFS Resource Sat-1 730km 56m 5 days Yes Yes large-scale deforestation IKONOS IKONOS – 11. canopy gaps.

GeoEye (resampled to 0. DigitalGlobe 1. 25 m quad pol standard roads.7m SWIR (8) logging vehicles PALSAR-2 PALSAR-JAXA varies 1-3 m Spotlight 15 days Varies Yes deforestation. Varies 0. canopy gaps. canopy gaps.5m multispectral 1-5 days No Yes Deforestation.5m) illegal fishing vessels / 1. 19 m dual pol No. canopy gaps. roads.31m panchromatic (1) < 1 day No Yes skid trails. 3-10 m Stripmap encroachment.GeoEye-1 GeoEye-1 .8m multispectral illegal fishing vessels / logging vehicles Radarsat-2 CSA varies 8 m quad pol fine 24 days varies Yes deforestation.4m panchromatic (1) Varies No Yes skid trails. log ponds 100 m wide ASAR ENVISAT-ESA varies 30 m polarization mode 36 days varies No deforestation 150 m Wide Swath mode 1 km Global Monitoring mode PALSAR PALSAR-JAXA varies 9 m Single pol 45 days Yes.24m multispectral (8) illegal fishing vessels / 3.quad pol roads.all mode No deforestation. log ponds REDD+ MRV MANUAL: CHAPTER 5. canopy gaps.1km 0.5m panchromatic Varies No Yes skid trails.65m multispectral (4) logging vehicles QuickBird QuickBird – Varies 0.6m (1) panchromatic Varies No Yes skid trails.0 – REMOTE SENSING OF LAND COVER CHANGE 119 . Constellation encroachment.4m multispectral (4) illegal fishing vessels / logging vehicles RapidEye BlackBridge 77km 6. DigitalGlobe 1. log ponds 30 m quad pol 100 m Scan SAR WorldView-3 WorldView – 13. DigitalGlobe 2. skid trails WorldView-2 WorldView-2 Varies 0. 100 m Scan SAR roads.

deforestation. in fires calibration REDD+ MRV MANUAL: CHAPTER 5.0 – REMOTE SENSING OF LAND COVER CHANGE 120 .Future Missions Sensor Satellite – Swath Resolution Repeat Cycle Systematic Operational Monitoring Agency Width Acquisitions Status Applications SPOT-7 Airbus 60km 1. Space 6. Defense & panchromatic (1) in calibration encroachment. then (2B) VIIRS NPV – NASA 3000km 750m 2 times per day Yes recently launched.0m roads.5m No Launched June 2014. log ponds multispectral (4) Sentinel-2 ESA 290km 10/20m/60m 5 day (once both 2A Anticipated launch deforestation A/B & 2A are launched) summer 2014 (2A).

cfm Quantum GIS http://www.aspx PCI Geomatics Biodiversity Informatics Facility INTERGRAPH: ERDAS Imagine SELECTED RESOURCES Online guides and other materials United Nations Space Science and Technology: American Museum of Natural History (AMNH) Biodiversity Informatics Random forests Software REDD+ MRV MANUAL: CHAPTER 5.php?section_id=33&content_id=138 GRASS GIS http://grass. Open source GIS and remote sensing software http://biodiversityinformatics.eo.clarklabs. http://www.intergraph.html Systems for World Surveillance.php?section_id=17 European Space Agency Earthnet Software EXELIS: ENVI http://www. Inc.0 – REMOTE SENSING OF LAND COVER CHANGE 121 .eu IDRISI GIS and Image Processing Software NASA Earth Observatory http://earthobservatory.

usgs. SAR Tool Kit ESA – polsarpro (Polarmetric SAR Data Processing and Educational Tool) Open source Alaska Satellite Facility .br/CDSR/ (Range of Landsat and CBERS REDD+ MRV MANUAL: CHAPTER NEST – Next ESA SAR toolbox (Range of data sources including the Landsat archive and selected imagery for a range of instruments including ASTER. and MODIS) RapidEye Catalog See5 classification software http://www. Ikonos.usgs.alaska.rapideye. Global Land Survey (GLS) Data access USGS Earth Resources Observation and Science Center (EROS) http://glovis.1 RAT – Radar Tools (Landsat Archive. as well as various MODIS products) Global Land Survey (GLS) 2005 products: Global Land-cover Facility http://www.0 – REMOTE SENSING OF LAND COVER CHANGE 122 . Orbview.htm Rulequest data mining tools.eo.inpe.Map Ready.html R statistical language as well as various ASTER and MODIS products) USGS LandsatLook Viewer http://landsatlook. (Enables searching of both LandsatLook images & Level1 Landsat data) National Research Institute (INPE) of Brazil http://www.

2012. Saatchi S. produced by the University of Maryland and distributed by Google. Proceedings of the National Academy of Sciences 14. Zutta. Mitchard.jpl.(Access to the BlackBridge RapidEye image archive) SPOT Catalog http://catalog.html REDD+ MRV MANUAL: CHAPTER NASA fundamentals of remote sensing http://gcmd. http://earthenginepartners. T.108(24):9899-904.T.L. Houghton.usgs. Baccini. These may be useful for national stratification of field sampling of biomass in a MRV system.nasa.html General remote sensing: Canada Centre of Remote Sensing Harris. M.doi. B. Petrova.1038/NCLIMATE1354 Pantropical National Level Carbon Stock Dataset http://www. N. Sun.ersdac. Lefsky. S. L. S J. remote-sensing/11740 USGS Change-tracking tool http://pubs. W. Brown.L. based on a suite of satellite data inputs. N. Benchmark map of forest carbon stocks in tropical regions across three continents. Hagen. M. Hackler.fas.nrcan.or. Two maps of global forest biomass.S.gc. Samanta and R. S. calibrated with plot data.spotimage. White. Morel. Beck.nasa. Salas. Dubayah. S.appspot. S. Laporte. Silman and A. with a 1-km resolution. D. M.aspx?language=UK (Access to the SPOT satellite archive) Earth Remote Sensing Data Analysis Center (ERSDAC) Tutorials The remote sensing tutorial: Federation of American Scientists (FAS) http://www. P. A. (Access to the ASTER imagery archive) Global tree-cover loss and biomass data Global Landsat-based estimates of tree-cover loss from 2000 to 2012. Nature Climate Change http://dx. S.0 – REMOTE SENSING OF LAND COVER CHANGE 123 . W. Lewis. Estimated carbon dioxide emissions from tropical deforestation improved by carbon-density maps. Buermann. 2011. R. M. http://carbon. Friedl. Walker. to remote sensing .Virtual Hawaii Radar polarimetry: Canada Centre of Remote Sensing http://www.pdf REDD+ MRV MANUAL: CHAPTER SAR-Land-Applications-Tutorial/sar_land_apps_1_theory.html An introduction to remote sensing CSIRO An introduction to radar remote sensing: Canada Centre of Remote Sensing ESA’s Synthetic Aperture radar: Land applications tutorial http://earth.esa.html NOAA’s Satellite and Information Service: Learning About Satellites and Remote Sensing ESA’s RADAR Tutorial http://earth.cmis.0 – REMOTE SENSING OF LAND COVER CHANGE 124

NW Washington.U. DC 20523 Tel: (202) 712-0000 Fax: (202) 216-3524 www.usaid. Agency for International Development 1300 Pennsylvania .S.