Data and Methods
Every year, the Global Forest Review (GFR) provides an independent assessment of the state of the world’s forests based on the best available geospatial data and analysis. A key distinguishing element of the GFR is its focus on insights derived from analysis of geospatial data and maps. In 2014, breakthroughs in global forest monitoring using satellite data, computer algorithms, and cloud computing resulted in the first global map of forest change at 30-meter resolution, depicting
The Global Forest Change data set, with its annual updates on tree cover loss and gain, provides a critical input to the report. The GFR also draws on spatial data and analysis techniques that are rapidly improving with the evolution of forest monitoring technologies and scientific methods. The report and the Data and Methods section will be updated annually to reflect the latest advances in data and data science.
Data and Methods is organized into three subsections: Data Sets, Methodology, and Indicators Overview. The Data Sets section describes the spatial data used in the GFR. The Methodology section describes techniques underpinning any calculations and analysis of the data conducted by World Resources Institute to derive and report results. The Indicators Overview section builds on these two sections by summarizing the data sets and methods used for each calculation.
The Global Forest Review (GFR) uses the best-available global spatial data on forests. Over 20 different global data sets come together to help us understand why our forests are changing and the impacts these changes have on people, climate, and biodiversity. Unless otherwise specified, the data descriptions below summarize definitions and methods outlined in published papers. Additional manipulation or processing of the data sets was not done for GFR analyses. The data sets are divided between the following types:
|Type||Data Set||Source||Spatial Resolution||Temporal Resolution||Years of Coverage||Spatial Coverage|
|Forest change||Tree cover loss||Hansen et al. (2013)||30-meter||Annual||2001-19||Global|
|Tree cover loss by dominant driver||Curtis et al. (2018)||10-kilometer||Annual||2001-19||Global|
|Tree cover gain||Hansen et al. (2013)||30-meter||12 years||2001-12||Global|
|Lower Mekong height and canopy||Potapov et al. (2019)||30-meter||Annual||2001-17||Lower Mekong|
|Tree cover change||Song et al. (2018)||5-kilometer||34 years||1982–2016||Global|
|Hot spots of primary forest loss||Harris et al. (2017)||Vector||17 years||2002–19||Tropics|
|Tree cover extent||Hansen et al. (2013)||30-meter||1 year||2000, 2010||Global|
|Primary forest||Turubanova et al. (2018)||30-meter||1 year||2001||Tropics|
|Intact forest landscapes||Potapov et al. (2017)||Vector||3 years||2000, 2013, 2016||Global|
|Tree plantations||Harris et al. (2019)||Vector||1 year||2015||Global|
|Mangroves||Bunting et al. (2018)||Vector||7 years||1996, 2007, 2008, 2009, 2010, 2015, 2016||Global|
|Commodities||Global cocoa, coffee, soy||MapSPAM||10-kilometer||1 year||2010||Global|
|Global pasture||Ramankutty et al. (2008)||10-kilometer||1 year||2000||Global|
|Brazilian pasture||Laboratório de Processamento de Imagens e Geoprocessamento (LAPIG)||30-meter||1 year||2018||Brazil|
|South America Soy||Song et al. (forthcoming)||30-meter||Annual||2001-18||South America|
|Oil palm, rubber, wood fiber||Harris et al. (2019)||Vector||1 year||2015||Select countries|
|Management||Protected areas||World Database on Protected Areas||Vector||1 year||Varies, see below||Select countries|
|Logging concessions||Varies, see below||Vector||1 year||Varies, see below||Select countries|
|Wood fiber concessions||Varies, see below||Vector||1 year||Varies, see below||Select countries|
|Oil palm concessions||Varies, see below||Vector||1 year||Varies, see below||Select countries|
|Roundtable on Sustainable Palm Oil (RSPO) concessions||RSPO member companies||Vector||1 year||Varies, see below||Select countries|
|Mining concessions||Varies, see below||Vector||1 year||Varies, see below||Select countries|
|Oil and gas concessions||Varies, see below||Vector||1 year||Varies, see below||Select countries|
|Carbon||Aboveground biomass density||Zarin and Woods Hole Research Center||30-meter||1 year||2000||Global|
|Water||Erosion risk||Qin et al. (2016)||10-kilometer||1 year||2015||Global|
|Urban watersheds||The Nature Conservancy||Vector||1 year||Unknown||Global|
|Social||LandMark||Global Platform of Indigenous and Community Lands||Vector||1 year||2019||Select countries|
|Population||Global Human Settlement Layer||250-meter||1 year||2015||Global|
|Conflict||Global Witness||Vector||Annual||2012-18||Select countries|
|Other||Ecozones||FAO (2012)||Vector||1 year||2010||Global|
|Peatlands||Indonesia Ministry of Agriculture||Vector||1 year||2012||Indonesia|
|Indonesian forest moratorium||Indonesia Ministry of Forestry and Environment||Vector||1 year||2019||Indonesia|
|Rural complex||Molinario et al. (2015)||30-meter||4 years||2000, 2005, 2010, 2015||Democratic Republic of the Congo|
|Countries||Database of Global Administrative Areas (GADM)||Vector||n/a||2019||Global|
Tree cover loss.
In this data set,
This data set has been updated seven times since its creation and now includes loss data up to 2019 (Version 1.7). The analysis method has been modified in numerous ways, including new data for the target year, reprocessed data for previous years (2011 and 2012 for the Version 1.1 update, 2012 and 2013 for the Version 1.2 update, and 2014 for the Version 1.3 update), and improved modeling and calibration. These modifications improve change detection for 2011–19, including better detection of boreal loss due to fire, smallholder rotation agriculture in tropical forests, selective logging, and short-cycle plantations. Eventually, a future “Version 2.0” will include reprocessing for 2000–10 data, but in the meantime, comparisons between the original 2001–10 data and the 2011–19 update should be performed with caution. Although we report on annual trends since 2001, we are more confident in the later part of the time series.
At the global scale, for the original 2001–12 product, the overall prevalence of false positives (detected as tree cover loss but, in reality, is not, also known as commission errors) in this data is 13 percent, and the prevalence of false negatives (not detected as tree cover loss but, in reality, is lost, also known as omission errors) is 12 percent, though the accuracy varies by biome and thus may be higher or lower in any particular location. The model often misses
Tree cover loss by dominant driver.
- Commodity-driven deforestation: Long-term, permanent conversion of forest and shrubland to a nonforest land use such as agriculture (including oil palm), mining, or energy infrastructure.
- Shifting agriculture: Small- to medium-scale forest and shrubland conversion for agriculture that is later abandoned and followed by subsequent forest regrowth.
- Forestry: Large-scale forestry operations occurring within managed forests and tree plantations.
- Wildfire: Large-scale forest loss resulting from the burning of forest vegetation with no visible human conversion or agricultural activity afterward.
- Urbanization: Forest and shrubland conversion for the expansion and intensification of existing urban centers.
For the purposes of statistics generated in the Global Forest Review, commodity-driven deforestation, urbanization, and shifting agriculture with primary forest are considered to represent permanent deforestation, whereas tree cover usually regrows in the other categories (forestry, wildfire, and shifting agriculture outside of
The data were generated using decision tree models to separate each 10-kilometer (km) grid cell into one of the five categories. The decision trees were created using 4,699 sample grid cells and use metrics derived from the following data sets: Hansen et al. (2013) tree cover, tree cover gain, and tree cover loss; National Aeronautics and Space Administration fires; global land cover; and population count. Separate decision trees were created for each driver and each region (North America, South America, Europe, Africa, Eurasia, Southeast Asia, Oceania), for a total of 35 decision trees. The final outputs were combined into a global map that is then overlaid with tree cover loss data to indicate the intensity of loss associated with each driver around the world.
Regional models were created, and training samples allowed for the interpretation of local land uses or management styles. A cell was categorized as commodity-driven deforestation if it contained clearings that showed signs of existing agriculture, pasture, or mining in the most recent imagery (after the tree cover loss occurred) as well as zero or minimal regrowth in subsequent years. Cells were categorized as shifting agriculture if the cell contained clearings that showed signs of existing agriculture or pasture in most recent imagery (after the tree cover loss occurred) as well as past clearings that contained visible forest or shrubland regrowth (gain) in historical imagery spanning 2001–15. Tree crops typically considered as agricultural commodities, such as oil palm, were classified accordingly as commodity-driven deforestation. The forestry class reflects a combination of wood fiber plantations and other forestry activity, including clear-cutting and selective cuts. Cells were categorized as wildfire when large swaths of fire scarring were visible in cleared areas, indicating that the loss event was driven by wildfire. The wildfire class excludes fire used to clear land for agriculture. Cells were categorized as urbanization if the loss of tree cover coincided with visible urban expansion or intensification.
This data set is intended for use at the global or regional scale, not for individual pixels. Individual grid cells may have more than one
Aside from commodity-driven deforestation, urbanization, and shifting agriculture with primary forest, which are assumed to represent permanent conversion from a forest to nonforest state, this data set does not indicate the stability or changing condition of the forest land use after the tree cover loss occurs. The data set does not distinguish between natural or anthropogenic wildfires, but it does distinguish fires for conversion or agricultural activity, which are not included in the wildfire class. Only direct drivers of forest disturbance are considered, and not indirect drivers such a demographic pressures or economic markets.
The accuracy of the data was assessed using a validation sample of 1,565 randomly selected grid cells. At the global scale, overall accuracy of the model was 89 percent, with individual class accuracies ranging from 55 percent (urbanization) to 94 percent (commodity-driven deforestation). The data has been updated since the original publication to include tree cover loss data from 2016 to 2019.
Tree cover gain.
Tree cover gain is defined as the establishment of tree canopy at the Landsat pixel scale in an area that previously had no tree cover but regrew to greater than 50 percent tree canopy cover density. Tree cover gain may indicate a number of potential activities, including natural forest growth or the crop rotation cycle of tree plantations.
Lower Mekong height and canopy.
Tree cover change.
Hot spots of primary forest loss.
The emerging hot spots analysis uses the annual Hansen et al. (2013) tree cover loss data set between the years 2002 and 2019, the Turubanova et al. (2018) primary forest extent data set for the year 2001, and the Esri ArcGIS Emerging Hot Spot Analysis geoprocessing tool. The tool uses a combination of two statistical measures: the Getis-Ord Gi* statistic to identify the location and degree of spatial clustering of forest loss and the Mann-Kendall trend test to evaluate the temporal trend over time. The analysis was run for individual countries, and its results are relative to the patterns and amount of loss in each country. It has been updated since the original publication to include the latest tree cover loss data.
Tree cover extent.
Intact forest landscapes.
To map IFL areas, the extent of forest areas was identified using greater than 20 percent tree canopy density in the Hansen et al. (2013)
Changes in the extent of IFLs were identified within the year 2000 IFL boundary using the global wall-to-wall Landsat image composite for 2016 and the global forest cover loss data set (Hansen et al. 2013). Areas identified as “reduction in extent” met the IFL criteria in 2000, but they no longer met the criteria in 2016. The main causes of change were clearing for agriculture and tree plantations, industrial activity such as logging and mining,
The world IFL map was created through visual interpretation of Landsat images by experts. The map may contain inaccuracies due to limitations in the spatial resolution of the imagery and lack of ancillary information about local land-use practices in some regions. In addition, the methodology assumes that fires near roads or other infrastructure may have been caused by humans and therefore constitute a form of anthropogenic disturbance. This assumption could result in an underestimation of IFL extent in the boreal biome.
The planted trees category in the SDPT includes forest
Approximately 15,000 Landsat scenes and 1,500 ALOS PALSAR (one-by-one degree) mosaic tiles were used to create optical and radar image composites covering the coastlines along the tropical and subtropical coastlines in the Americas, Africa, Asia, and Oceania.
The classification was confined using a mangrove habitat mask, which defined regions where mangrove ecosystems can be expected to exist. The mangrove habitat definition was based on geographical parameters such as latitude, elevation, and distance from ocean water. Training for the habitat mask and classification of the 2010 mangrove mask was based on randomly sampling 38 million points using the mangrove masks (for the year 2000) of Giri et al. (2011)
The Landsat 7 scan-line error affects the classification in certain areas, resulting in striping artifacts in the data. Classification accuracy was assessed with over 53,800 randomly sampled points across 20 randomly selected regions. Overall accuracy was 95.25 percent, and the user’s and producer’s accuracies for the mangrove class were estimated at 97.5 percent and 94.0 percent, respectively. Factors such as satellite data availability (due to clouds, cloud shadows, and Landsat 7 scan-line error), mangrove species composition, and level of degradation can all lead to local variations in accuracy. The mangrove seaward border is generally also more accurately defined than the landward side, where distinction between mangrove and certain terrestrial vegetation species can be unclear.
Areas known to be missing in this version (2.0) of the data set include Bermuda (United Kingdom); Europa Island and the Wallis and Futuna Islands (France); Fiji, east of longitude 180° east; Guam and Saipan (United States); Kiribati; Maldives; and Peru, south of latitude 4° south.
Global cocoa, coffee, soy.
South America soy.
Oil palm, rubber, wood fiber.
Logging concessions. Managed forests refers to areas allocated by a government for harvesting timber and other wood products in a public
This data set is assembled by aggregating data for multiple countries. Source and date information can be found in the table below.
Logging concession data sources and dates
|Cameroon||Ministry of Forestry and Wildlife and World Resources Institute (WRI)||Unknown|
|Canada||Global Forest Watch Canada||2016|
|Central African Republic||Ministry of Water, Forests, Hunting, and Fishing and WRI||Unknown|
|Democratic Republic of the Congo||Ministry of the Environment, Nature Conservation, and Tourism and WRI||Unknown|
|Equatorial Guinea||Ministry of Agriculture and Forests and WRI||2013|
|Gabon||Ministry of Forest Economy, Water, Fisheries, and Aquaculture and WRI||Unknown|
|Indonesia||Ministry of Environment and Forestry||2018|
|Sarawak, Malaysia||Earthsight and Global Witness||2010|
|Republic of the Congo||Ministry of Forest Economy and Sustainable Development and WRI||2013|
Wood fiber concessions. Wood fiber concession refers to an area allocated by a government or other body for establishing fast-growing tree plantations for the production of timber and wood pulp for paper and paper products. This data set is assembled by aggregating data for multiple countries. Source and date information can be found in the table below.
Wood fiber concession data sources and dates
|Indonesia||Ministry of Environment and Forestry, Asia Pulp & Paper, and Asia Pacific Resources International Limited||2018|
|Republic of the Congo||Ministry of Agriculture and WRI||Unknown|
|Sarawak and Sabah, Malaysia||Earthsight and Global Witness||2011|
Oil palm concessions. Oil palm concession refers to an area allocated by a government or other body for industrial-scale oil palm plantations. This data set is assembled by aggregating data for multiple countries. Source and date information can be found in the table below.
Oil palm concession data sources and dates
|Indonesia||Ministry of Environment and Forestry||2012|
|Sarawak, Malaysia||Sarawak Dayak Iban Association (SADIA), Aidenvironment, and Earthsight||2010|
|Republic of the Congo||Ministry of Agriculture and WRI||2013|
Roundtable on Sustainable Palm Oil (RSPO) concessions. This data set maps the concession boundaries of RSPO member companies through December 2017, including both certified and noncertified concessions, as well as concessions where the certification status is unknown. The concession boundaries were provided to the RSPO by member companies.
Mining concessions. Mining concession refers to an area allocated by a government or other body for the extraction of minerals. The terminology for these areas varies from country to country. Concession is used as a general term for licenses, permits, or other contracts that confer rights to private companies to manage and extract minerals from public lands; terminology varies at the national level, however, and includes mineral or mining permits, tenures, licenses, and other terms. This data set is assembled by aggregating data for multiple countries. Source and date information can be found in the table below.
Mining concession data sources and dates
|Brazil||National Department of Mineral Production||2019|
|Cameroon||Ministry of Mines, Industry and Technological Development||Unknown|
|Cambodia||Open Development Cambodia||2014|
|Canada||Global Forest Watch Canada||2016|
|Colombia||Tierra Minanda; Agencia Nacional de Minería de Colombia||2008|
|Democratic Republic of the Congo||Ministry of Mines Mining Registry||Unknown|
|Gabon||Ministry of Mines, Petroleum, and Hydrocarbons and WRI||Unknown|
|Mexico||Secretaría de Economía||2015|
|Peru||Instituto Geológico Minero y Metalúrgico (INGEMMET)||2015|
|Republic of the Congo||Ministry of Mines and Geology and WRI||Unknown|
Oil and gas concessions. Oil and gas concession refers to an area allocated by the government to companies who explore for and produce oil, natural gas, and other hydrocarbons. The terminology for these areas varies from country to country. Concession is used as a general term for licenses, permits, or other contracts that confer rights to private companies to manage and extract oil and natural gas from public lands; terminology varies at the national level, however, and includes permits, tenures, licenses, and other terms. This data set is assembled by aggregating data for multiple countries. Source and date information can be found in the table below.
Oil and gas concession data sources and dates
|Argentina||Ministry of Energy and Mining||2015|
|Brazil||National Agency of Petroleum, Natural Gas and Biofuels||2017|
|Colombia||National Agency of Hydrocarbons||2017|
|Ecuador||Ministry of Non-renewable Natural Resources||2013|
|Peru||Petróleos del Perú S.A.||2018|
Aboveground biomass density.
To produce this population density and distribution data set, researchers mapped global built-up areas, which are defined as all aboveground constructions intended for human or animal sheltering or to produce economic goods. The locations of these built-up areas were established using Landsat imagery analysis. An additional source used to compile this data set was the Gridded Population of the World (GPW) data set assembled by Columbia University’s Center for International Earth Science Information Network. The GPW data set consists of census population data and bolstered the built-up areas data by enabling researchers to estimate residential population. To present this data as grid cells, GPW data was disaggregated from census or administrative units.
Overall, the GHSL data set is an accurate and high-resolution estimate of global population. Known issues with this data include the insufficient availability of global test sets with the right scale, time period, and reliability to validate and improve the GHSL. Another known challenge is the lack of remote sensing studies that compare the use of different sensors to detect human settlements.
Indonesian forest moratorium.
The data set was created by characterizing forest clearing using spatial models in a geographical information system, applying morphological image processing to the Central African Forests Remotely Assessed (Forets d'Afrique Central Evaluee par Teledetection; FACET) product. This process allowed for the creation of maps for 2000, 2005, 2010, and 2015, classifying the rural complex and previously homogenous
All analyses in the Global Forest Review draw heavily on per-pixel geodesic area calculations for accurate global area estimations of forests. This means that the precise geodesic area of each 30-meter (m) pixel across the globe is calculated and then summed for each year of loss and unique area of interest, such as countries or protected areas. Due to distortions from projecting the three-dimensional surface of the earth onto a flat surface, the area of a 30 m pixel can vary from roughly 900 m2 at the equator to roughly 200 m2 at the poles. These area differences are accounted for when using geodesic area calculations. Unless otherwise specified, all calculations are run at a 30 percent tree canopy density threshold as of the year 2000.
- Area calculation: Sum the geodesic area of all pixels within an area of interest.
- Extent calculation: Sum the geodesic area of all pixels within the
tree coverextent raster data set.
- Tree cover loss calculation: Sum the geodesic area of all
tree cover losspixels within an area of interest (AOI; e.g., country boundaries or protected areas).
- Rate of loss calculation: Loss area in current year minus loss area in past year divided by loss area in past year. Only countries with at least 100,000
hectaresof tree cover in the year 2000 were included.
- Percent of loss calculation: Divide loss of current year by earlier
- Carbon storage calculation: The aboveground biomass density data set is formatted as biomass per hectare. To convert values to carbon per pixel, each biomass pixel is multiplied by the geodesic area (in hectares) of that pixel to get biomass per pixel, and then divided by 0.5 to convert biomass to carbon. Finally, sum the aboveground biomass pixel values that overlap with the tree cover extent raster data set.
- Carbon emissions calculation: The aboveground biomass density data set is formatted as biomass per hectare. To convert values to carbon per pixel, each biomass pixel is multiplied by the geodesic area (in hectares) of that pixel to get biomass per pixel, and then divided by 0.5 to convert biomass to carbon. Finally, sum the aboveground biomass pixel values that overlap with the tree cover loss data set. All of the aboveground biomass is considered to be “committed” emissions to the atmosphere upon clearing. Emissions are “gross” rather than “net” estimates, meaning that information about the fate of land after clearing, and its associated carbon value, is not incorporated. Emissions associated with other carbon pools, such as soil carbon, are not included.
Using the above data sets and methodologies, the Global Forest Review (GFR) assesses the state of the world’s forests and provides insight into how they are changing year to year based on 18 indicators. The next section outlines each statistic produced by GFR authors, along with the data set and method summary used to generate each calculation.
Indicators of Forest Change
- Forest Extent
- Forest Loss
- Primary Forest Loss
- Deforestation Linked to Agriculture
- Forest Gain
- Trees outside Forests
Indicators of Forest Condition
Indicators of Forest Designation
Indicators of Biodiversity and Ecological Services
Indicators of Social and Governance Issues
The Forest Extent Indicator aims to monitor the total area of
|In 2010, the world had 3,929 million hectares (Mha) of tree cover covering 30 percent of land on Earth.||Tree cover extent||Extent calculation on 2010 tree cover extent|
|In the tropics, mapped tree plantations account for roughly 2 percent of all tree cover; the remaining 98 percent of tree cover can be assumed to be natural or seminatural forest.||Tree plantations||Extent calculation on tree plantations; includes all area within the tree plantation data set, divided by tree cover extent in the tropics|
|Tropical and subtropical forests account for 58 percent of 2010 tree cover by area. Boreal forests make up 27 percent of tree cover. Temperate forests account for about 15 percent of tree cover.||Ecozones; tree cover extent||Extent calculation on 2010 tree cover extent by ecozone|
|Primary forests account for roughly 50 percent of all forests in the tropics (1,030 Mha). Brazil, the Democratic Republic of the Congo, and Indonesia have the most primary forest in absolute terms, whereas French Guiana, Gabon, and Suriname have the highest proportion of their total land area covered by primary forest.||Ecozones; primary forest||Extent calculation on primary forest divided by extent calculation on 2010 tree cover extent by ecozone|
This indicator aims to monitor the total area of
Tree cover loss calculation in
|The world has lost 386 million hectares (Mha) of tree cover since the turn of the century, equivalent to about 10 percent of global tree cover in 2001.||Tree cover loss; tree cover extent||Tree cover loss calculation in 2000 extent|
|Tree cover loss has been rising in recent history, from 13.4 Mha of tree cover loss in 2001 to 24.2 Mha in 2019.||Tree cover loss||Tree cover loss calculation|
|Forestry is associated with 115 Mha of tree cover loss. . . Commodity-driven deforestation is associated with 98 Mha . . . Wildlife is associated with 86 Mha. . . Shifting agriculture is associated with 80 Mha . . . Urbanization is associated with 3 Mha.||Tree cover loss; tree cover loss by dominant driver||Tree cover loss in all drivers of tree cover loss categories|
|Roughly one-third of tree cover loss since 2000 was likely to be deforestation, characterized by permanent loss of forest.||Tree cover loss; tree cover loss by dominant driver||Tree cover loss calculation in shifting agriculture overlapping primary forest, commodity-driven deforestation, and urbanization driver categories|
|The remaining two-thirds of tree cover loss was likely more temporary in nature.||Tree cover loss; tree cover loss by dominant driver||Tree cover loss calculation in wildfire, forestry, and nonprimary shifting agriculture driver classes|
|Though only half (192 Mha) of global tree cover loss this century occurred in the tropical ecozones, the tropics accounted for more than 95 percent of all global deforestation.||Tree cover loss; tree cover loss by dominant driver; ecozones||Tree cover loss calculation in tropical ecozones and shifting agriculture overlapping primary forest, commodity-driven deforestation, and urbanization driver categories|
|The annual rate of tropical tree cover loss nearly doubled from 6.7 Mha in 2001 to 11.9 Mha in 2019.||Tree cover loss; ecozones||Tree cover loss calculation in tropical ecozones|
|Although Brazil and Indonesia have experienced the largest area of tree cover loss this century, these two countries have become less dominant as the rate of tree cover loss in all other tropical countries combined has increased by 154 percent since 2001.||Countries; tree cover loss||Tree cover loss and rate of loss calculation in countries|
|Temperate and boreal forests have experienced 148 Mha of gross tree cover loss since the turn of the century. . .||Tree cover loss; ecozones||Tree cover loss calculation in temperate and boreal ecozones|
|. . . with almost 98 percent related to temporary factors (forestry and wildfire).||Tree cover loss; tree cover loss by dominant driver||Tree cover loss calculation in forestry and wildfire driver categories|
|Canada, Russia, and the United States together make up 94 percent of all tree cover loss related to wildfire and 54 percent of all loss related to forestry.||Countries; tree cover loss; tree cover loss by dominant driver||Tree cover loss calculation in countries and forestry and wildfire driver categories|
|The United States lost more than 2.8 Mha of tree cover to urbanization between 2001 and 2019, about 25 times more than Canada.||Countries; tree cover loss; tree cover loss by dominant driver||Tree cover loss calculation in countries and urbanization driver category|
This indicator aims to monitor the total area of
|The humid tropics have lost 60 million hectares (Mha) of primary forests since the turn of the century, representing 5.9 percent of their extent in 2001.||Tree cover loss; primary forest||Tree cover loss and percent loss calculation in primary forest|
|Just three countries—Brazil, the Democratic Republic of the Congo, and Indonesia—accounted for 65 percent of this loss.||Countries; tree cover loss; tree cover loss by dominant driver||Tree cover loss calculation in primary forest and tree cover loss by dominant driver|
|Roughly two-thirds of this loss was related to conversion for commodity production (industrial-scale agriculture, mining, oil and gas, etc.), with an additional one-third related to
||Tree cover loss; primary forest; tree cover loss by dominant driver||Tree cover loss calculation in primary forest and tree cover loss by dominant driver|
|Russia and Canada experienced the highest levels of
||Countries; tree cover loss; ecozones; intact forest landscapes||Tree cover loss calculation in primary forest and intact forest landscapes|
This indicator estimates the role of specific agricultural commodities in agriculture-linked
Maps of croplands and
To produce these calculations, the commodity extent maps are overlaid with
This analysis cannot determine whether areas were deforested for the purpose of growing the specific commodity or for other reasons (e.g., harvesting wood products). These data also do not measure possible leakage effects of these commodities on deforestation—for example, where expansion of a commodity displaces other forms of farming that contribute to deforestation elsewhere. As such, this indicator refers to forests that are “replaced by” specific commodities rather than deforestation “driven by” specific commodities.
The data sets and methods used rely on a number of assumptions and have a number of associated caveats, which are further described in Goldman et al. (2020).
This indicator aims to monitor the total land area that has transitioned from an unforested to forested state each year.
Due to data limitations, the indicator currently measures
|The world experienced 80.6 million hectares (Mha) of tree cover gain between 2001 and 2012.||Tree cover gain||Tree cover gain calculation|
|More than half (55 percent) of tree cover gain between 2001 and 2012 occurred in tropical and subtropical ecozones.||Tree cover gain; ecozones||tropical and subtropical ecozones|
|Around a fifth of all gain (21 percent) in this part of the world occurred in existing tree plantations and presumably does not reflect regeneration of natural forests.||Tree cover gain; tree plantations||Tree cover gain calculation in tree plantations|
|The remaining 45 percent of tree cover gain between 2001 and 2012 occurred in temperate and boreal ecozones.||Tree cover gain; ecozones||Tree cover gain calculation in boreal and temperate ecozones|
The Trees outside Forests Indicator aims to monitor trees that are growing outside
This indicator aims to monitor the area of forest worldwide that has experienced degradation due to natural or human causes. Because it is not currently possible to measure forest degradation directly at a global scale, this indicator uses disturbed forest area that has not necessarily been completely cleared as a proxy for where degradation may have occurred, acknowledging that not all forest disturbance results in forest degradation.
The indicator measures disturbed areas using three approaches that are currently possible using medium-resolution satellite imagery. The three measures are not mutually exclusive (i.e., areas of different kinds of disturbance often overlap) and they are therefore estimated individually rather than summed:
- Forest area experiencing a partial (more than 20 percent and less than 90 percent) loss of tree canopy cover (as measured at 250-meter resolution via satellite imagery)
- Tree cover extent experiencing tree cover loss due to fire
- Intact forest landscapes (defined as forests that are undisturbed by human activity and larger than 50,000 hectares) that can no longer be considered intact due to evidence of human disturbance
|Worldwide, 86 million hectares (Mha) of tree cover loss were associated with fire between 2001 and 2019, affecting 2.1 percent of global tree cover area.||Tree cover loss; tree cover loss by dominant driver||Tree cover loss calculation in wildfire drivers of tree cover loss category|
|Worldwide, 120 Mha of forest area that were considered intact in 2000 could no longer be considered intact in 2016, corresponding to a reduction of approximately 7.5 Mha of intact forest per year. . .||Intact forest landscapes (IFLs)||Subtract extent calculation in 2016 IFLs from 2000 IFLs|
|. . . and a total reduction of 9 percent of intact forest area.||IFLs||Subtract extent calculation in 2016 from 2000 IFLs and divide by extent calculation in 2000 IFLs|
|Russia experienced the largest reduction of intact forest area (29 Mha), primarily due to fire.||Countries; IFLs||Subtract extent calculation in 2016 IFLs from 2000 IFLs by country|
|Romania saw the largest percentage decline, with its last remaining tract of intact forest fragmented by new transport infrastructure.||Countries; IFLs||Subtract extent calculation in 2016 IFLs from 2000 IFLs by country, divided by 2000 IFLs|
|Paraguay also experienced a notable decline, with an 80 percent decrease in forest area that could be considered intact between 2000 and 2016 due to the clearing of the Chaco for cattle ranching.||Countries; IFLs||Subtract extent calculation in 2016 IFLs from 2000 IFLs by country, divided by 2000 IFLs|
The Forest Recovery Indicator aims to monitor the area of
This indicator aims to monitor the extent of
|Approximately 20 percent of global forest area is currently under some form of legal protection.||Tree cover extent; protected areas||Extent calculation in 2010 tree cover extent and protected areas|
|Existing protected areas cover 36 percent of tropical primary forest and 28 percent of global intact forest landscapes.||Primary forest; intact forest landscapes (IFLs); protected areas||Extent calculation in primary forest, IFLs, and protected areas|
|In 2019, protected areas experienced 3.4 million hectares (Mha) of tree cover loss, including 0.62 Mha of primary forests and 0.69 Mha of intact forests.||Tree cover loss; primary forest; IFLs||Tree cover loss calculation in primary forest, IFLs, and protected areas|
|A total of 40.1 Mha of tree cover loss has occurred within protected areas since 2001, and a total of 4.7 Mha of tree cover gain also occurred in these areas between 2001 and 2012.||Tree cover loss; tree cover gain; protected areas||Tree cover loss and gain calculation in protected areas|
|In strict nature reserves, wilderness areas, and national parks (a subset of all protected areas that often have the most importance for biodiversity conservation), tree cover loss has increased since 2001 by an average of 18 percent per year, and total tree cover extent has shrunk by 3.9 percent.||Tree cover loss; tree cover extent; protected areas||Tree cover loss and extent calculation in Category I and II protected areas|
The Production Forests Indicator aims to monitor the extent and location of
Due to a lack of comprehensive spatial data showing the location of production forests worldwide, this indicator cites nongeospatial statistical data from the Global Forest Resources Assessment 2020 to provide estimates of the global extent of production forests.
To approximate the location of production forests, the indicator uses the tree cover loss by dominant driver data set produced by Curtis et al. (2018),
To estimate the location of tree plantations, the indicator uses data from the Spatial Database of Planted Trees by Harris et al. (2019). This database aggregates maps of tree plantations and
|In these nine countries, managed
||Managed forest concessions; intact forest landscapes; tree cover extent||Tree cover extent calculation in intact forest landscapes, managed forest concessions, and countries|
|…which are estimated to cover roughly 75 percent (173 million
||Tree plantations||Area calculation in tree plantations|
This indicator aims to monitor the amount of carbon stored in
|Tropical rainforests account for only 30 percent of global tree cover but contain 50 percent of the world’s carbon stored in trees.||Tree cover extent; aboveground biomass density; ecozones||Carbon storage calculation in tropical ecozones|
This indicator aims to monitor the amount of carbon dioxide (CO2) released into or absorbed from the atmosphere. As
At this time, Global Forest Watch monitors CO2 emissions related to annual
|It is now possible to map annual gross emissions from tropical tree cover loss. These emissions have increased significantly since the turn of the century, from 2.4 gigatons (Gt) of CO2 in 2001 to 4.0 Gt of CO2 in 2019.||Tree cover loss; aboveground biomass density; ecozones||Carbon emissions calculation in tropical ecozones|
|Emissions from tree cover loss within tropical
||Tree cover loss; aboveground biomass density; primary forest||Carbon emissions calculation in primary forest|
|Brazil, the Democratic Republic of the Congo, and Indonesia were responsible for two-thirds of total CO2 emissions from tropical primary forest loss between 2001 and 2019.||Countries; tree cover loss; aboveground biomass density; primary forest||Carbon emissions calculation in primary forest, by country|
This indicator aims to monitor the extent of
- Forests in areas with high erosion risk, determined by an area’s precipitation, elevation, slope, soil properties, and land cover
- Forests within watersheds that supply water to the world’s urban areas
- Mangrove forests
|As of 2010, . . . 47 percent of the world’s tree cover was located in areas defined as being at medium to high risk of erosion.||Tree cover extent; erosion risk||Tree cover extent calculation in 2010 tree cover extent and high erosion risk|
|In 2019, 10.6 million
||Tree cover loss; erosion risk||Tree cover loss and rate of loss calculation in high erosion risk|
|Australia, Brazil, and Indonesia accounted for 30 percent of global tree cover loss in such areas in 2019.||Countries; tree cover loss; erosion risk||Tree cover loss calculation in high erosion risk, by country|
|Since 2001, 90 Mha of tree cover have been lost in watersheds that supply the world’s urban areas—representing an 8.6 percent reduction in tree cover in these watersheds since 2000.||Tree cover loss; tree cover extent; watersheds
||Tree cover loss calculation in urban watersheds|
|The Beaumont, Monroe, Port Arthur, and Wilmington watersheds in the United States; the Belém watershed in Brazil; and the Concepción watershed in Chile have lost 30 percent or more of their tree cover since 2000.||Tree cover loss; tree cover extent; watersheds
||Tree cover loss calculation in urban watersheds|
|Since 2016, 124 Mha of tree cover loss has occurred within mangrove forests. Over 80 percent of this loss has occurred in Brazil, Cuba, Indonesia, Malaysia, and the United States.||Countries; tree cover loss; mangroves||Tree cover loss calculation in 2016 mangroves, by country|
|As of 2016, coastal mangrove forests covered 13.3 Mha across 106 countries, representing a tiny fraction of global tree cover—just 0.3 percent.||Tree cover extent; mangroves||Area extent calculation in mangroves; total mangrove area was included in estimation and divided by 2010 tree cover extent|
|One-fifth of coastal mangroves are located in Indonesia, and another quarter are located in Australia, Brazil, Mexico, and Nigeria.||Countries; mangroves||Area extent calculation in mangroves, by country|
The Indigenous and Community Forests Indicator measures changes in the extent of forestlands that are both legally titled and customarily held by Indigenous Peoples and other
|The LandMark data set includes approximately 136,000 indigenous and community land maps and indicative areas, representing about 12 percent of the world’s land.||LandMark||Area extent calculation in LandMark|
|These areas contain approximately 16 percent of the world’s
||LandMark; IFLs; carbon density||Area extent and carbon storage calculation in LandMark and IFLs|
|Between 2013 and 2018, the percentage of
||Tree cover loss; LandMark; countries||Tree cover loss calculation in LandMark and countries|
|When comparing tree cover loss across all indigenous and community lands between two time periods (three-year averages between 2013–15 and 2016–18), tree cover loss was 1.6 times higher in indigenous and community lands that were not acknowledged by government versus those that were.||Tree cover loss; LandMark||Tree cover loss calculation in LandMark|
The At-Risk Populations Indicator measures the number of people who are potentially vulnerable to losing sources of food and other
This indicator provides a general estimate of at-risk populations, but some populations in these areas will not be at risk. There are also some populations outside these areas who are at risk, but they are not included in this estimate. No error estimates are provided.
|Approximately 136 million people currently live within five kilometers of an emerging hot spot of forest loss.||Hot spots of primary forest loss; population||Sum population within hot spots|
|The Democratic Republic of the Congo, India, Cameroon, Vietnam and Brazil have the highest number of people living within deforestation hot spots, whereas French Guiana, Gabon, the Solomon Islands, Equatorial Guinea, and the Central African Republic have the highest proportion of their total population close to these hot spots.||Hot spots of primary forest loss; population||Sum population within hot spots, by country|
The Forestland-Related Conflict Indicator measures the amount of social conflict and violence stemming from decisions about the use of forestlands and resources. As competition for land and natural resources expands and Indigenous Peoples and local communities act to protect
The Global Forest Change data offer an annual view of the world’s
- Not all tree cover is a forest. Satellite data are effective for monitoring changes in
tree cover, but forests are typically defined as a combination of tree cover and land use. For example, agricultural tree cover, such as an oil palm plantation, is not usually considered to be forest. As such, satellite-based monitoring systems may overestimate forest area unless combined with additional land-use data sets. No land-use data set currently exists at an adequate resolution or update frequency to enable this analysis at global scale.
- Not all tree cover loss is deforestation. Defined as permanent conversion of forested land to other land uses,
deforestationcan only be identified at the moment trees are removed if it is known how the land will be used afterward. In the absence of a global data set on land use, it is not possible to accurately classify tree cover lossas permanent (i.e., deforestation) or temporary (e.g., where it is associated with wildfire, timber harvesting rotations, or shifting cultivation) at the time it occurs. However, new models analyzing spatial and temporal trends in tree cover loss are enabling better insights into the drivers of loss.
- Tree cover is a one-dimensional measure of a forest. Many qualities of a forest cannot be measured as a function of tree cover and are difficult, if not impossible, to detect from space using existing technologies. Forests that are vastly different in terms of form and function—such as an
intact primary forestand a planted forestmanaged for timber production—are nearly indistinguishable in satellite imagery based on tree cover. Detecting forest degradationthrough remote sensing is also challenging because degradation often entails small changes occurring beneath the forest canopy.
- Tree cover gain is more difficult to measure than loss. Whereas tree cover loss is distinctly visible at a specific moment in time,
tree cover gainis a gradual process and is thus more difficult to discern from one satellite image to the next. Annual reporting of tree cover loss has not been matched by annual reporting of tree cover gain, resulting in an unbalanced view of global forest change dynamics. Ongoing improvements in detection methodologies are likely to deliver the first annualized global tree cover gain data set by 2020.
- Tree cover loss and gain do not equal net forest. Due to variation in research methodology and date of content, tree cover, tree cover loss, and tree cover gain data sets cannot be compared accurately against each other. Accordingly, “net” loss cannot be calculated by subtracting figures for tree cover gain from tree cover loss, and current (post-2000) tree cover cannot be determined by subtracting figures for annual tree cover loss from year 2000 tree cover.
Definitions for basic forest-related concepts and terms are often varied and controversial. For consistency and clarify, the Global Forest Review (GFR) uses the following definitions across the report:
agricultural tree crop: Trees cultivated for their food, cultural, or economic values. These include oil palm, rubber, cocoa, cashew, mango, oranges (citrus), plantain, banana, and coconut.
biodiversity intactness: The proportion and abundance of a location's original forest community (number of species and individuals) that remain.
biodiversity significance: The importance of an area for the persistence of forest-dependent species based on range rarity.
deforestation: The change from forest to another land cover or land use, such as forest to plantation or forest to urban area.
forest: Forests include tree cover greater than 30 percent tree canopy density and greater than 5 meters in height as mapped at a 30-meter Landsat pixel scale.
forest concession: A legal agreement allowing an entity the right to manage a public forest for production purposes.
forest degradation: The reduction in a forest’s ability to perform ecosystem services, such as carbon storage and water regulation, due to natural and anthropogenic changes.
forest disturbance: A discrete event that changes the structure of a forest ecosystem.
forest fragmentation: The breaking of large, contiguous forests into smaller pieces, with other land cover types interspersed.
forest management plan: A plan that documents the stewardship and use of forests and other wooded land to meet environmental, economic, social, and cultural objectives. Such plans are typically implemented by companies in forest concessions.
hectare: One hectare equals 100 square meters, 2.47 acres, or 0.01 square kilometers and is about the size of a rugby field. A football pitch is slightly smaller than a hectare (pitches are between 0.62 and 0.82 hectares).
intact forest: A forest that contains no signs of human activity or habitat fragmentation as determined by remote sensing images and is large enough to maintain all native biological biodiversity.
land and environmental defenders: People who peacefully promote and protect rights related to land and/or the environment.
loss driver: The direct cause of forest disturbance.
low tree canopy density: Less than 30 percent tree canopy density.
managed forest concession: Areas where governments have given rights to private companies to harvest timber and other wood products from natural forests on public lands.
mosaic restoration: Restoration that integrates trees into mixed-use landscapes, such as agricultural lands and settlements, where trees can support people through improved water quality, increased soil fertility, and other ecosystem services. This type of restoration is more likely in deforested or degraded forest landscapes with moderate population density (10–100 people per square kilometer).
natural forest: A forest that is grown without human intervention.
persistent loss and gain: Forests that have experienced one loss or one gain event from 2001 to 2016.
plantations: An area in which trees have been planted, generally for commercial purposes.
planted forest: A forest composed of trees that have been deliberately planted and/or seeded by humans.
primary forest: Old-growth forests that are typically high in carbon stock and rich in biodiversity. The GFR uses a humid tropical primary rainforest data set, representing forests in the humid tropics that have not been cleared in recent years.
production forest: A forest where the primary management objective is to produce timber, pulp, fuelwood, and/or nonwood forest products.
seminatural forest: A managed forest modified by humans, which can have a different species composition from surrounding natural forests.
shifting agriculture: Temporary loss or permanent deforestation due to small- and medium-scale agriculture.
tree cover: All vegetation greater than five meters in height and may take the form of natural forests or plantations across a range of canopy densities. Unless otherwise specified, the GFR uses greater than 30 percent tree canopy density for calculations.
tree cover gain: The establishment of tree canopy in an area that previously had no tree cover. Tree cover gain may indicate a number of potential activities, including natural forest growth or the crop rotation cycle of tree plantations.
tree cover loss: The removal or mortality of tree cover, which can be due to a variety of factors, including mechanical harvesting, fire, disease, or storm damage. As such, loss does not equate to deforestation.
trees outside forests: Trees found in urban areas, alongside roads, or within agricultural land are often referred to as Trees Outside Forests (TOF).
tree plantation: An agricultural plantation of fast-growing tree species on short rotations for the production of timber, pulp, or fruit.