Chapter 47. Global mapping

ABSTRACT

FRA 2000 developed new global forest and ecological maps which give spatial definition to area statistics of the survey findings from individual countries and regions, providing a synoptic view of worldwide forest cover. The global ecological zoning map provides an important means of aggregating global information on forests or other natural resources according to their ecological character. Together the maps are useful for the analysis and depiction of worldwide forest cover according to the forests' ecological character.

The forest cover map was developed using coarse-resolution satellite imagery. In previous global assessments, the means and technology did not exist to produce a global map based on satellite imagery. Thus the technical map based on state-of-the-art technology replaces mere illustrations of global forests.

The ecological zoning map, based on a standard global classification, was produced using existing national and regional potential vegetation maps, climate data and satellite imagery.

A third map of protected forests was also developed and used in estimating the area of forest under formal protection worldwide. Inputs were collected from countries around the world.

Each map is generated from a corresponding computerized geographic information system (GIS) database, which makes it possible to combine the maps with different spatial and statistical data, permitting new perspectives on the world's forests. Computerized maps and databases are more easily updated than conventional maps, and they set the groundwork for future assessments as well. Digital versions of the maps are available to researchers and the general public through the FRA Web site (www.fao.org/forestry/fo/fra/index.jsp).

INTRODUCTION

FRA 2000 produced three global maps: a forest map, an ecological map and a protected areas map. Each map is generated from a corresponding computerized geographic information system (GIS) database. This makes it possible to combine the maps with spatial and statistical data from other sources for computation of statistics at the global, regional and ecological zone levels, permitting new perspectives on the world's forests. Computerized maps and databases are more easily updated than conventional maps and lay the groundwork for future assessments.

The global forest map shows the extent and location of major forest formations throughout the world (Figure 47-1). The ecological map can be combined with other maps or data to help quantify or depict global forests according to their ecological character (Figure 47-2). The protected area map depicts the location, extent and type of protected area for each country of the world.

The forest and ecological zone maps are useful to a scale of 1:40 000 000, although enlargements are possible up to 1:10 000 000. Digital versions of the maps are available to researchers and the general public through the FRA Web site, with the exception of the protected areas map, which is managed exclusively by the UNEP World Conservation Monitoring Centre (UNEP-WCMC).

FAO worked with several cooperators in the development of the various maps, including the EROS Data Center (EDC), United States; UNEP-WCMC, United Kingdom; the International Institute for Applied Systems Analysis (IIASA), Austria; the Ecological Laboratory of Toulouse (LET), France; the Tropical Science Center, Costa Rica; the Autonomous University of Mexico (UNAM); the Canadian Forest Service (CFS); the Forest Service of the United States Department of Agriculture (USDA); Damascus University, Syrian Arab Republic; the Institute of Remote Sensing Applications (IRSA), China; the Chinese Academy of Sciences; and the Australian Bureau of Rural Sciences (BRS). EDC conducted all the image processing for the forest map and the global mosaic for the ecological zoning map. UNEP-WCMC compiled the global protected areas map. FAO organized and coordinated the work and carried out final quality control and edge-matching for all the maps. Other partners contributed valuable assistance to the technical construction and thematic content of the maps.

Figure 47-1. FRA 2000 global forest cover map

Figure 47-2. FRA 2000 global ecological zone map

Table 47-1. FRA 2000 global land cover map legend, definitions and representative land cover types

FRA 2000 class	FAO definition	Representative land cover
Closed forest	Land covered by trees with a canopy cover of more than 40 percent and height exceeding 5 m. Includes natural forests and forest plantations.	Tropical/subtropical moist forest Temperate broadleaf mixed forest Subtropical/temperate conifer plantation Boreal conifer forest
Open or fragmented forest	Land covered by trees with a canopy cover between 10 and 40 percent and height exceeding 5 m (open forest), or mosaics of forest and non-forest land (fragmented forest). Includes natural forests and forest plantations.	Northern boreal/taiga open conifer or mixed forest Southern Africa woodland Tropical fragmented/degraded forest
Other wooded land	Land either with a 5 to 10 percent canopy cover of trees exceeding 5 m height, or with a shrub or bush cover of more than 10 percent and height less than 5 m.	Mediterranean closed shrubland Tropical woody savannah
Other land cover	All other land, including grassland, agricultural land, barren land, urban areas.	Grassland, cropland, non-woody wetland, desert, urban
Water	Inland water.	Inland water

The development of the three global maps represented a major technical challenge for FRA 2000. Each was produced using the best available information for the purpose. The forest cover map was developed using coarse-resolution satellite imagery, the ecological map from national and regional potential vegetation maps and climate data, and the protected areas map from independent maps and point information supplied by countries.

The global maps provide a synoptic overview of the worldwide situation regarding forests, ecological zones and protected areas. They were used in conjunction with statistical data for the FRA 2000 reports on forest area by ecological zone, forest area under protection, protected areas within ecological zones and other parameters.

FOREST COVER MAP

The FRA 2000 forest map took three years to complete and shows the location and distribution of forests according to FRA 2000 classification criteria.

Overall consistency at the global level was viewed as an important objective for all the maps. For forest cover mapping, this could only be achieved by using a common input source such as satellite imagery and applying similar classification criteria for all areas. The classification scheme for the map was developed using the same criteria that were used in FAO's global assessment based on country statistical data and in its high-resolution remote sensing sampling programme (Table 47-1). Consequently, the global map could be integrated into the overall FRA framework and can be used in conjunction with other data sets. The map can also be used as a simple visual aid to show the location and extent of forests around the world, according to FAO's terminology.

One of the difficult and expensive tasks in producing a global map from satellite data, including imagery from the Advanced Very High Resolution Radiometer (AVHRR), is piecing together a large amount of data to produce a single cloud-free data set. Because clouds obscure forests, as well as other land cover, they must be eliminated before mapping can begin. Therefore, the forest map relied to a large extent on the Global Land Cover Characteristics Database (GLCCD) produced by EDC. This database proved invaluable to the mapping, since many of the problems of cloud cover and reflectance anomalies had been resolved through preprocessing and the use of a multi-date composite containing only the best image data. Nevertheless, the lack of good imagery prevented the mapping of several Pacific Islands. Source data for the forest map were drawn from the 1995-1996 data set, which was the latest imagery available in the GLCCD archive. This imagery consisted of five calibrated AVHRR bands and a Normalized Difference Vegetation Index (NDVI) band (Zhu and Waller 2001).

Although correspondence between the GLCCD and FRA classification schemes was considered generally good, the entire range of GLCCD classes could not be recoded easily into FRA 2000 classes. For example, in South America 34 of the original 167 classes required further processing for FAO's map (Zhu et al. 1999). Consequently, EDC adopted a methodology based on a combination mixture model with scaling of NDVI values and the visible band based on pixel positions along the infrared band. Regional variations in forest cover and associated reflectance required stratification of the processing into geographic divisions and adjustment of the models according to their respective conditions (Zhu and Waller 2001).

Once an advanced draft forest map was developed in 1999, EDC worked with FAO and UNEP-WCMC on validation and quality control. UNEP-WCMC used its extensive map archive to identify areas that needed further processing, and FAO sent copies to experts and FAO field offices around the world for feedback. This input was used over the final year to refine the map before assessment of its accuracy.

Determining the map's accuracy was viewed as an important step in the mapping exercise, since the final map was going to be used for technical work in conjunction with other data sets. EDC employed the use of an existing set of validation points from the International Geosphere and Biosphere Programme (IGBP) and full land-cover data sets available from the United States and Chinese Governments (Zhu and Waller 2001). FAO also conducted an evaluation of the map using 117 interpreted thematic mapper scenes from the tropics. The results of all these evaluations showed that the average accuracy of the map for all forest classes is about 80 percent. The closed forests are more accurately mapped than the average, and the open and fragmented forests are somewhat less accurately mapped. Other wooded lands are the least accurately mapped of the three classes.

The global forest cover map provides spatial definition for the area statistics and survey findings from the individual countries and regions. In previous global assessments, the means and technology did not exist to produce a global map based on satellite imagery. Thus, FRA 2000 has replaced artists' depictions of global forests with a technically correct map based on state-of-the-art technology.

For illustrative purposes the map was reproduced in Robinson Projection. However, because it exists in a GIS format, it is possible to transform the map (or portions of the map) into other projections according to specific requirements.

Forest area estimates were not derived from the map. However, the map served as a spatial framework for the integration of country statistics which were then used in conjunction with the ecological zoning and protected areas maps to estimate the fraction of forests under protection and in the various ecological zones.

ECOLOGICAL ZONE MAP

The underlying strategy for the FRA ecological zoning reflected both the thematic and technical needs of the map as well as the many operational constraints that were expected in its development. In terms of ecosystem principles, the map requirements were such that zones or classes were defined and mapped using a holistic approach. That is, both biotic and abiotic components of ecosystems were considered in the zoning scheme. Beyond the thematic content and zoning, practical aspects of digital cartographic production, such as data availability, currency, scale and the associated reliability of the map inputs, were also taken into account (Simons 2001).

FAO conducted two preliminary studies to identify specific alternatives and constraints in the development of a global ecological zone (GEZ) map appropriate for FRA 2000 purposes (Preto 1998; Zhu 1997). Findings from these studies, experience in the development of the tropical ecological zone map for FRA 1990, and recommendations from other parties consulted in the process indicated that FAO could not complete an entirely new global ecological zoning map by 2000 because of the large amount of scientific, organizational and financial resources and time required. FAO therefore focused on identifying an existing scheme that might be used or adapted to the programme's needs. A Workshop on Global Ecological Zones Mapping, held in Cambridge, United Kingdom in July 1999, and attended by experts from 15 countries, helped set the framework.

Because of the enormity of conducting the work on a global scale, a classification scheme had to be chosen that would meet FAO's thematic requirements, be practical to construct with available resources and meet the scrutiny of diverse users from all parts of the world. Existing schemes were each developed for specific purposes according to various environmental criteria. Macroclimate (temperature and precipitation) was an element used by most (Preto 1998; WCMC 1992). Since macroclimate correlates well with the potential vegetation associated with a particular locale, it was considered a logical basis for the FRA ecological zoning as well.

Table 47-2. Ecological zone breakdown used in FRA 2000

EZ Level 1 - Domain		EZ Level 2 - Global Ecological Zone
Name	Criteria (equivalent to Köppen-Trewartha climatic groups)	Name (reflecting dominant zonala vegetation)	Code	Criteria (approximate equivalent of Köppen-Trewartha climatic types, in combination with vegetation physiognomy, and one orographic zone within each domain)
Tropical	All months without frost: in marine areas over 18°C	Tropical rain forest	TAr	Wet: 0-3 months dry,b during winter
		Tropical moist deciduous forest	TAwa	Wet/dry: 3-5 months dry, during winter
		Tropical dry forest	TAwb	Dry/wet: 5-8 months dry, during winter
		Tropical shrubland	TBSh	Semi-arid: evaporation > precipitation
		Tropical desert	TBWh	Arid: all months dry
		Tropical mountain systems	TM	Approximately > 1 000 m altitude (local variations)
Subtropical	Eight months or more over 10°C	Subtropical humid forest	SCf	Humid: no dry season
		Subtropical dry forest	SCs	Seasonally dry: winter rains, dry summer
		Subtropical steppe	SBSh	Semi-arid: evaporation > precipitation
		Subtropical desert	SBWh	Arid: all months dry
		Subtropical mountain systems	SM	Approximately > 800-1000 m altitude
Temperate	Four to eight months over 10°C	Temperate oceanic forest	TeDo	Oceanic climate: coldest month over 0°C
		Temperate continental forest	TeDc	Continental climate: coldest month under 0°C
		Temperate steppe	TeBSk	Semi-arid: evaporation > precipitation
		Temperate desert	TeBWk	Arid: All months dry
		Temperate mountain systems	TM	Approximately > 800 m altitude
Boreal	Up to three months over 10°C	Boreal coniferous forest	Ba	Vegetation physiognomy: coniferous dense forest dominant
		Boreal tundra woodland	Bb	Vegetation physiognomy: woodland and sparse forest dominant
		Boreal mountain systems	BM	Approximately > 600 m altitude
Polar	All months below 10°C	Polar	P	Same as domain level

^a Zonal vegetation: resulting from the variation in environmental, i.e. climatic, conditions in a north-south direction.

^b A dry month is defined as the month in which the total precipitation expressed in millimetres is equal to or less than twice the mean temperature in degrees Centigrade.

For the choice of climatic parameters to be used in the FRA 2000 map a number of global systems were surveyed (Köppen 1931; Trewartha 1968; Thornthwaite 1933; Holdridge 1947). Köppen modified by Trewartha was selected as the best candidate because of the number of classes that corresponded well to FRA 2000 needs. Moreover, while Köppen-Trewartha is based on climate, there is a demonstrated good correspondence between its subzones or climatic types and the natural climax vegetation types and soils within them (Bailey 1996).[56]

FAO, in cooperation with EDC and UNEP-WCMC, thus developed a prototype zoning scheme for FRA 2000 based on Köppen-Trewartha. The zoning was made hierarchical using Köppen-Trewartha's climatic groups and climatic types as FAO ecological zone levels 1 and 2, respectively (Table 47-2). A third level was also tested during the pilot project, representing the differentiation within the first two levels according to landform - distinguishing mountains with altitudinal zonation from lowland plains. This third level was ultimately not used.

At level 1, the broadest level, equivalent to Köppen-Trewartha's climatic groups, five domains are distinguished based on temperature: tropical, subtropical, temperate, boreal, polar.

At the second level, 20 classes or ecological zones are distinguished, which indicate broad zones of relatively homogeneous vegetation, such as tropical rain forest, tropical dry forest and boreal coniferous forest. The names of the global ecological zones reflect the dominant zonal vegetation. Typical azonal vegetation types, for instance mangroves, heath and swamps, are not separately classified and mapped.

Level 2 is the reference or working level for the GEZ mapping. The ecological zones were delineated by using both macroclimate data and existing climax or potential vegetation maps. Use of vegetation maps ensured a more precise delineation of the ecological zones. If generalized climate maps had been used alone, the zones of the final map would probably have corresponded poorly to boundaries of homogeneous vegetation transitions.

Within each domain (level 1) a zone of mountain systems is distinguished at level 2. Mountain systems usually contain a variety of vegetation types and include forests, alpine shrubs, meadows and bare rock. The current global framework cannot address the high, mostly small-scale diversity of mountain habitats. The polar domain is not further subdivided, as it is treeless, and only very sparse shrub or grass vegetation occurs locally. Here the second level is equivalent to the first.

A main principle in delineating the global ecological zones involves aggregating or matching regional ecological or potential vegetation maps into the global framework. The following steps can be distinguished:

• identification of Köppen-Trewartha climatic types and mountains occurring in a region to approximate the level 2 ecological zone class of the FAO scheme;
• establishment of correspondence between regional/national potential vegetation types and the global ecological zones;
• final definition and delineation of the global ecological zones, using the maps and source data consulted in the first two steps;
• edge-matching between adjacent maps;
• validation.

To ensure the best use of regional knowledge and information, existing regional/national maps on vegetation, biogeography, ecology and climate were used to generate the GEZ map. In some countries, such as the United States, classification is based on the Köppen-Trewartha climate system and translation to the FAO scheme was straightforward. In other cases, a more thorough study of mapping criteria, including physiognomy, phenology, floristics and dynamics of vegetation types, was needed to establish the correspondence. A benefit of using the existing country/regional maps is that they could form the basis or provide supporting information for more detailed regional ecological zoning beyond FRA 2000 (see Table 47-3).

The country/regional vegetation maps also helped in harmonization of ecological zone boundaries across countries or regions. The experts who attended the Cambridge workshop contributed in a major way to definition of the ecological zones of their respective regions as well as to edge-matching between adjoining geographic regions.

Both the existing FRA 1990 ecofloristic zone map and several existing regional maps were produced using the ESRI Arc/Info GIS software. Thus, it was convenient for the rest of the work to be conducted using Arc/Info, or at least to be Arc/Info importable. After study of the digital map in the Arc/Info coverage environment and confirmation that the digital version had appropriate attributes for the ecological zones (represented in the map by polygons), the coverage was edited and attributes for FAO ecological zone levels 1 and 2 were added.

Two problems occurred in polygon edge-matching along country and regional boundaries. One was mismatch of polygon definition translations between polygons in adjacent maps. This problem was generally easy to solve by going back to the original maps, checking the translation and modifying as needed. The other problem was the misalignment of lines of the polygons on both sides, even though they may have had the same labels. To resolve this problem, FAO manually edited the coverage and changed the locations of the boundaries. This sometimes required verification using ancillary data and maps such as composites of United States National Oceanic and Atmospheric Administration (NOAA) AVHRR spectral bands, classified continental-scale land cover data (such as the United States Geological Survey [USGS] global land cover database) and digital elevation model (DEM) data.

Table 47-3. Source maps used for the delineation of FAO global ecological zones

Region	Name of map	Scale	Projection	Thematic information / classification criteria
Canada and Mexico	Ecological regions of North America (CEC 1997)	million	Lambert Azimuthal Equal Area	Holistic classification system based on climate, soils, landform, vegetation and also land use. Hierarchical system: 15 Level I ecological regions and 52 Level II regions.
United States	Ecoregions of the United States (Bailey 1995)	million	Lambert Azimuthal Equal Area	Classification based on Köppen climate system: broad domains equivalent to climate groups, subdivided into divisions approximately equivalent to climate types.
Central America	National Holdridge life zone maps, transformed to a regional base map (Bolanos & Watson 1991; De la Cruz 1976; Hartshorn 1984; Holdridge 1962; Holdridge & Tosi 1971; Tosi 1970; Tosi & Hartshorn 1978)	Various scales Base map at 1:1.5 million	x	Holdrige life zones are defined using the parameters (bio)temperature, rainfall and evapotranspiration.
South America, Africa, Tropical Asia	Ecofloristic zones maps (LET 2000)	million	Lat-Long	groups of ecofloristic zones are defined, based on climate, vegetation physiognomy and physiography, i.e. altitude. The EFZ identifies the most detailed ecological units, based on the additional criteria of flora and geographic location.
Near East	Vegetation map of the Mediterranean zone (UNESCO/FAO 1970)	million	x	Distribution of potential vegetation formations in relation to climate. The various formations are distinguished mainly on the basis of physiognomy.
Europe	General map of the natural vegetation of Europe (Bohn et al. 2000)	million	Equidistant_Conic	Distribution of potential natural plant communities corresponding to the actual climate and edaphic conditions. At broadest level 19 vegetation formations defined, of which 14 zonal and 5 azonal formations.
Former USSR	Vegetation map of the USSR (Isachenko et al. 1990)	million	Lambert Azimuthal Equal Area	Distribution of broad vegetation formations related to climate, altitude and also current land use. 133 vegetation classes are aggregated into 13 categories of vegetation.
China	Geographic distribution of China's main forests (Zhu 1992)	x	x	Main aim to identify and map China's forest vegetation. A hierarchical classification is used based on climate and distribution of forest types and tree species. 27 forest divisions are mapped.
Australia	Interim biogeographic regionalisation for Australia (Thackway & Cresswell 1995)	million	Albers Equal Area	Major attributes to define biogeographic regions are: climate, lithology/geology, landform, vegetation, flora and fauna and land use. A total of 80 IBRA regions have been mapped.
Caribbean, Mongolia, Korean Peninsula, Japan, New Zealand, Pacific Islands	Terrestrial ecoregions of the world (WWF 2000)	x	Lat-Long	Ecoregions are defined by shared ecological features, climate and plant and animal communities. Main use is for biodiversity conservation.

Following the classification and guidelines outlined above, the global map was compiled in a region-by-region approach. Case studies on North America and South America provided useful experiences and guidelines for GEZ mapping in other regions. In the course of the work regional experts actively participated or were consulted. EDC was responsible for producing the ecological zone maps for the temperate and boreal regions and jointly with FAO compiled the global map and database, while LET, Toulouse produced the ecological zone maps for the tropical regions, i.e. South America, Africa and Asia. FAO provided overall technical and conceptual guidance. After the Cambridge meeting in July 1999, it took one year to produce a draft global map. The draft map was reviewed at a meeting in Salt Lake City, Utah, United States (5-7 July 2000), and the final map and database were completed by October 2000.

Table 47-4. Distribution of forests by ecological zone, 2000

Ecological zone	Total forest %	Africa %	Asia %	Oceania %	Europe %	North and Central America %	South America %
Tropical rain forest	28	24	17	-	-	1	58
Tropical moist deciduous	11	40	14	6	-	9	31
Tropical dry	5	39	23	-	-	6	33
Tropical mountain	4	11	29	-	-	30	30
Total tropical forests	47	28	18	1	-	5	47
Subtropical humid forest	4		52	8	-	34	6
Subtropical dry forest	1	16	11	22	30	6	14
Subtropical mountain	3	1	47	-	13	38	1
Total subtropical forests	9	2	42	7	7	37	5
Temperate oceanic forest	1	-	-	33	33	9	25
Temperate continental forest	7	-	13	-	40	46	-
Temperate mountains	3	-	26	5	40	29	-
Total temperate forests	11	-	17	4	39	39	2
Boreal coniferous forest	19	-	2	-	74	24	-
Boreal tundra woodland	3	-	-	-	19	81	-
Boreal mountain	11	-	1	-	63	36	-
Total boreal forests	33	-	2	-	65	34	-
Total forests	100	17	14	5	27	14	23

After production of the regional GEZ maps the global GEZ map was composed from all the regional tiles. Edge-matching was an issue, particularly for the vast area of Europe and Asia, where a number of different tiles had to be brought together with large bordering areas. The delineation of ecological zones between bordering areas of Europe and the former Union of Socialist Soviet Republics (USSR) matched well, with only small adjustments needed. The same applies to the ecological zone boundaries between Europe and the Near East. More work was needed to match the tiles for tropical Asia, China and the former USSR; the task was complicated by the presence of extensive mountain systems on the border areas. After the edge-matching problems were resolved, the regional tiles were registered to a global base map, ESRI's Digital chart of the world, 1st edition, December 1994 (base scale 1:1 000 000). The GEZ map, together with other global maps produced by FRA 2000, is presented on the FAO Forestry Web site (www.fao.org/forestry/fo/fra/index.jsp) under "Global maps".

The GEZ map can be used to aggregate information on forest resources by ecological zone. Consequently, it is now possible to produce reports according to the natural characteristics of the vegetation rather than by national boundaries, which frequently cut across natural ecosystems. This is particularly important today, with the growing awareness that many environmental problems are not national in character. For example, analysis of global change in climate and forest resources and of change in regional biological corridors requires information with a broad geographical context. Through ecological zone mapping, valuable insight is being obtained about the characteristics of forest resources which may serve to identify and resolve issues of importance to many countries, entire regions or the planet as a whole. For FRA 2000 reporting purposes, an overlay of the forest cover map with the ecological zoning map was used to derive area statistics on forests according to ecological zones (Table 47-4).

PROTECTED AREAS MAP

UNEP-WCMC served as the lead collaborator in mapping protected areas and was wholly responsible for compiling the information. UNEP-WCMC maintains a database for protected areas around the world and worked in cooperation with FAO to update this information for FRA 2000 under a formal Letter of Agreement.

A draft protected areas map for each country was circulated to over 200 countries in 1997 and 1998. The maps depicted the location and boundaries of previously registered protected areas. The maps were accompanied by a survey form to facilitate the information collection. About 25 percent of the countries responded to the survey and provided new information to UNEP-WCMC. After determining its suitability, UNEP-WCMC digitized the data and entered them into the geographic information system. Some follow-up with the countries - with little resulting additions to the information base - was carried out through May 1999, when the activity was formally closed (UNEP-WCMC 2000).

Table 47-5. International and national data for protected areas

Region	Polygons			Points
	National	International	Total	National	International	Total
Africa	1 926	293	2 219	2 088	74	2 162
Asia	3 907	288	4 195	2 384	107	2 491
Europe	2 1468	1 587	23 055	19 478	1 915	21 393
North and Central America	10 119	352	10 471	4 722	92	4 814
Oceania	816	427	1 243	2 739	53	2 792
South America	2 436	158	2 594	1 413	48	1 461
Antarctica			0	28		28
Other	25	12	37	3 156	517	3 673
Total	40 697	3 117	43 814	36 008	2 806	38 814

Information entered into the database was classified according to two aggregations of the six IUCN categories (categories Ia to II and III to VI) and digitized as either points or polygons. The attribute data included the IUCN designation and various metadata needed to understand the source and currency of the information.

The protected areas map contains the latest and best overall compilation of spatial information on the world's protected areas. The database consists of over 43 000 polygons and 38 000 points representing over 55 000 national and international protected areas (Table 47-5). UNEP-WCMC will continue to update the information as a core part of its programme.

The protected areas map was used in FRA 2000 to estimate the status of the protection of forests around the world.

CONCLUSIONS

The global forest map is a useful visual aid for perceiving the location and extent of the major forest areas of the world (although FAO country statistics are still derived through other means). In addition, the map can be used as an overlay to combine it with the protected areas map to show areas of protected forest or with the GEZ map to show forest distribution by ecological zone. Produced using advanced image processing techniques and satellite imagery, it is the first map of its kind for a global assessment.

The GEZ map provides an important means of aggregating global information on forests and other natural resources according to their ecological characteristics. It is the only global tool of its nature, in the sense that it has been compiled and reviewed by a body of experts through an international process, is based on technical input from around the world and is digital, geometrically corrected and registered to a map base. The map provides an important tool for all users conducting global studies with ecological parameters. This is especially important, as the use of global ecological zoning is expected to grow in importance with the increasing need for information relating to climate change (Kyoto Protocol), desertification and biological diversity conservation. The map will also continue to be important for FAO's periodic global assessments.

The protected areas map shows the worldwide location and distribution of protected areas according to UNEP-WCMC data. In conjunction with statistical and spatial data on forests, the map can be used to estimate the amount of forest at present under some sort of protection. UNEP-WCMC plans to update the map regularly. Illustrative examples of protected areas are available on the FAO Web site.

Thanks to FRA 2000, future efforts in mapping of global forest cover, ecological zoning and protected areas now have a sound basis on which to build. Because the information is digital and geometrically corrected to a geographic map base, new data can be relatively easily integrated with existing information. The forest cover and ecological zoning maps are available to users around the world at no charge on the Internet. FAO hopes that the newly available information will be useful to other global change projects and scientific endeavours.

BIBLIOGRAPHY

Bailey, R.G. 1989. Explanatory supplement to ecoregions of the continents. Environmental Conservation, 16(4).

Bailey, R.G. 1995. Description of ecoregions of United States. USDA Forest Service Publication No. 1391, Washington, DC.

Bailey, R.G. 1996. Ecosystem geography. New York, Springer Verlag.

Bailey, R.G. 1998. Ecoregion map of North America. USDA Forest Service Publication No. 1548, Washington, DC.

Bohn, U., Gollub, G. & Hettwer, C. 2000. General map of the natural vegetation of Europe. Scale 1:10 million. Bonn, Germany, Federal Agency for Nature Conservation.

Bolanos, R. & Watson, V. 1991. Mapa ecológico de Costa Rica. Scale 1:200 000. San Jose, Costa Rica, Tropical Science Center.

Commission for Environmental Cooperation (CEC). 1997. Ecological regions of North America. Montreal, Canada.

De la Cruz, R. 1976. Mapa de zonas de vida de Guatemala. Scale 1:500 000. Instituto Nacional Forestal (INAFOR), Ministerio de Agricultura, Guatemala.

Ecological Laboratory of Toulouse (LET). 2000. Ecofloristic zones and global ecological zoning of Africa, South America and tropical Asia. Prepared for FAO-FRA 2000 by M.F. Bellan. Toulouse, France.

Hartshorn, G. 1984. Ecological life zones of Belize. Scale 1:1 400 000. In Belize country environmental profile: a field study. San Jose, Costa Rica, Trejos Hnos. Suc.

Holdridge, L.R. 1947. Determination of world plant formations from simple climatic data. Science, 105: 367-368.

Holdridge, L.R. 1962. Mapa ecológico de Honduras. Scale 1:1 000 000. Organization of American States.

Holdridge, L.R. & Tosi, J.A. 1971. Mapa ecológico de la República de Nicaragua. Scale 1:500 000.

Isachenko, T.I., Karamysheva, Z.V., Ladygina, G.M. & Safronova, I.N. 1990. Map of vegetation of the USSR. Scale 1:4 million. Moscow, Institute of Geography, RAS. (in Russian)

Köppen. 1931. Grundrisse der Klimakunde. Berlin, Walter de Gruyter Co.

Preto, G. 1998. A proposal for the preparation of the global eco-floristic map for FRA 2000. Rome, FAO. (unpublished)

Simons, H. 2001. Global ecological zones mapping. FRA Working Paper No. 56. Rome, FAO.

Thackway, R. & Cresswell, I.D. (eds). 1995. An interim biogeographic regionalisation for Australia: a framework for setting priorities in the National Reserves system cooperative program. Version 4.0. Canberra, Australia, Australian Nature Conservation Agency.

Thornthwaite, C.W. 1933. The climates of Earth. Geographic Review, 23.

Tosi, J.A. 1970. Mapa ecológico de Panamá. Scale 1:500 000. Proyecto de Inventario y Demostraciones Forestales. Panama/UNDP/FAO.

Tosi, J.A. & Hartshorn, G.S. 1978. Mapa ecológico de El Salvador: sistema de zonas de vida del Dr. L. R. Holdridge. Scale 1:300 000. Ministerio de Agricultura y Ganadería de El Salvador/Centro Agronómico Tropical de Investigación y Enseñanza, Subprograma de Suelos Análogos de Centro América.

Trewartha, G.T. 1968. An introduction to weather and climate. New York, McGraw-Hill.

UNEP-WCMC. 2000. Global FRA 2000 final report. United Kingdom. (unpublished)

UNESCO/FAO. 1970. Vegetation map of the Mediterranean zone. Explanatory notes. Arid Zone Research Series No. 30.

WCMC. 1992. Global biodiversity: status of the earth's living resources. London, Chapman & Hall.

WWF. 2000. Terrestrial ecoregions of the world. Washington, DC.

Zhu, Z. 1992. Geographic distribution of China's main forests. Nanjing, China, Nanjing Forestry University.

Zhu, Z. 1997. Develop a new global ecological zone map for GFRA 2000. Rome, FAO.

Zhu, Z. & Waller, E. 2001. Global forest cover mapping for the United Nations Food and Agriculture Organization Forest Resources Assessment 2000 Program. Project Report to FAO. Sioux Falls, South Dakota, USA, EROS Data Center.

Zhu, Z., Waller, D., Davis, R. & Lorenzini, M. 1999. Global forest cover map. Interim Progress Report. FRA Working Paper No. 19. Rome, FAO.