The below methodology framework is based on pilot tests and expert consultations. It is conceptually well underpinned, but is continually subject to discussion.
The Global Forest Survey is designed as a field inventory with multi-stage sampling to determine an efficient set of sites for the field work. This means that the surveyed parameters are measured or investigated in the field, using established measurement and interview techniques. The statistical design will ensure that aggregations of field data to larger units is possible.
The basic design identifies the country as a primary sampling unit, one purpose being to develop a project package for implementation by country. Although the GFS will not in itself be designed to produce national estimates, the country is a logical unit for implementation, suitable also for bilateral agreements. The implementation of a GFS package may also be combined with more intensive surveys for national purposes, or other, broader forestry projects.
The primary sampling unit is the country (or a group of countries). The reason for using administrative boundaries are (a) to facilitate national capacity building in the implementation, (b) to be able to tie in with existing national survey grids, (c) to enable project formulation for the field component on a bilateral basis.
Within the primary sampling unit a systematic grid of sites is laid out. The tentative size of the sites is 1 km2 – a size that was also used in the pilot field survey in Thailand (Dalsgaard et al. 2000). It is anticipated that a grid with 500-5000 sites will be needed for a country (Figure 2). A global grid will be developed for use in countries where a national grid is missing.
Figure 2. Schematic illustration of the primary sample grid for a country. The grid should tie in to existing national survey grids. It is tentatively estimated that 500 - 5000 sites will be needed depending on the size of the country.
From the systematic grid of sites above, a smaller number (say 50-500) will be sampled for field work. This sampling must ensure that the field work becomes efficient with respect to the prioritized forestry variables. All sites will therefore first be surveyed using mainly remote sensing techniques. The purpose is to make a stratification of the sites for the field sampling (Figure 3). Based on the stratification, a field sample will be created (Figure 4). The field sample should cover all strata, but with varying sampling intensity. If and where relevant, the field sample will be considered permanent and would be revisited within 5-10 years.
Figure 3. The survey of the primary sampling grid is made to stratify the material for field sampling. One stratification variable could be the relative extent of forest cover within the site. The top graph below is an illustration of how the sites may be distributed as to the extent of forest cover for two different countries, one heavily forested ("Sweden") and one with predominantly desert ("Algeria"). The bottom graph illustrates how strata (I-IV) may be assigned based on the distribution, and given varying sampling intensity (0.01 - 10%).
Figure 4. Illustration of sites sampled for field work (dark dots). Sampling is made after stratification of the primary grid. It is tentatively estimated that 50-500 sites/country will be needed to produce regional estimates of basic parameters.
Sample field sites will visited and surveyed in detail with respect to forestry, environmental and socio-economic parameters. A core set of global parameters will be collected on all sites. Parameters of local or regional interest may also be identified. National institutions and professionals will have a key role in the final design and implementation of the field work.
The range of global parameters to be surveyed will be broad and is currently under investigation. One starting point that is considered is the systematic standard of products and services done by the International working group on taxonomic databases for plant sciences (Cook 1995) that provides a standardized classification scheme of benefits (Table 6), that can be expanded and customized for forestry purposes. The survey can then be designed to indentify which benefits that are relevant for the field site in question, and then quantify the supply (resources) and demand (including production) for these benefits, including information needed to project these factors over time. In addition to resources and production information, socio-economic parameters would be surveyed, for example land ownership, forest/land management practices, local/national policies that apply. The survey would be done through a combination of measurements, observations and interviews, using established methodologies.
Table 6. Level one classification of plant uses according to Cook (1995).
1 |
Food |
2 |
Food additives |
3 |
Animal food |
4 |
Bee plants |
5 |
Invertebrate food |
6 |
Materials (including eg wood) |
7 |
Fuels |
8 |
Social uses |
9 |
Vertebrate poisons |
10 |
Non-vertebrate poisons |
11 |
Medicines |
12 |
Environmental uses |
13 |
Gene sources |
A field pilot study has been conducted in Thailand in April-May 2000. The objectives were to test some basic technologies and methods for the development of the GFS. The findings were positive. Collaboration with national institutions and local stakeholders was fruitful. GPS and computers was of great assistance. Experiences for further development of a field manual for the GFS were gained (Dalsgaard et al. 2000, Figure 5).
Figure 5. Field site (1 km2)from pilot GFS Survey in Thailand (Dalsgaard et al. 2000). Aerial photograph with major land use pattern (left) and photo from the on-site field work (right).
All data related to the GFS will be delivered to and managed by FAO. Data are made available to the country and, in restricted form, to wider audiences. Analyses on regional and global levels will be done by FAO and published as part of the global assessments.
FAO directly manages the sampling and stratification in close collaboration with the concerned countries. FAO also develops definitions, guidlines and manuals for the field component.
The succesful implementation of the GFS will depend on close collaboration with and involvement by national institutions and professionals. The GFS concept implies a strong focus on delivery of quality results, rather than conventional capacity building activities. It is anticipated that the request for delivery and quality adds a positive edge to the capacity building.
Capacity building will occur in all phases of the country-based survey, from the initial sampling design, to the field work and reporting. Capacity building will primarily be in the form of project management, on-the-job training, information management and analyses, and reporting. The network of professionals involved in the GFS will become a strong platform for further national or international initiatives.
Ten phases are identified over a period of nine years according to Table 7 and Figure 6.
Table 7. Identified phases/activities and expected outputs
PHASE |
Expected results |
1. Methodology development |
(Ongoing) Preliminary guidelines for the different phases of the GFS. Reports from expert consultations and pilot tests |
2. Pilot country surveys |
Survey results from appr. 5 countries, based on the preliminary guidelines. |
3. Design consolidation |
Ratification by FAO member countries and agreement on implementation plan |
4. Implementation, led by FAO |
Survey results from 10% of world's countries. Implementation led by FAO in collaboration with the countries. |
5. Implementation, by countries |
Survey results from 30% of world's countries for which the implementation is done by the countries themselves (in liaison with FAO) |
6. Implementation, bilateral |
Survey results from 60% of world's countries. Implementation in countries supported by bilateral or other arrangement (in liaison with FAO). |
7. Information system development |
Web based information system for storage and analysis of GFS data, integrated with FAO's corporate systems. |
8. Information system maintenance |
Continuous improvement of information system and maintenance of data. |
9. Biannual reporting of results |
GFS results to be published biannually. |
10. Secretariat and data management (FAO) |
Coordination of project formulations and communication of survey data. |
Figure 6. Time frame for identified phases
