Annexes (continued)

ANNEX 9
(Section 3.1)
List of Sampling Units and Satellite Images

ANNEX 10
(Section 3.2.1)
Regional Workshops
Summary of Conclusions and Recommendations

TCDC WORKSHOP ON METHODOLOGY FOR DEFORESTATION ASSESSMENT IN SOUTH EAST ASIA

BANGKOK 6–17 MAY 1991

SUMMARY

Funding Agencies
Unit of Technical Cooperation among Developing Countries (TCDC), United Nations Development Program

Implementing Agencies
Royal Forest Department of Thailand and FAO Regional Office for Asia and the Pacific (RAPA), within the framework of the Forest Resources Assessment 1990 Project, with the former also as host institution.

Participants
Seventeen national resource specialists from nine countries: India, Indonesia, Laos, Malaysia, Myanmar, Papua New Guinea, Philippines, Thailand, Vietnam.

Location
Royal Forest Department, Remote Sensing and Forest Mapping Division, Bangkok, Thailand.

CONCLUSIONS AND RECOMMENDATIONS

Classification

1. The Workshop underscored that FAO's classification of woody vegetation cover for the interpretation of high resolution satellite data is based on physiognomic aspects in order to assess the changes of forest area as well as the changes of forest density.
2. The Workshop agreed that there is a need for having a common classification structure that would make the results produced in various countries compatible and would allow consistent synthesis at national. sub-regional, regional and global levels.
3. The Workshop noted that it is necessary and appropriate for a country to reserve the freedom and flexibility to further sub-divide the proposed classes into more specific classes in order to meet the requirements at national level. Such further sub-division could include specific forest formations and floristic groups such as: coniferous and broadleaved forest; mangrove forest; bamboo forest; etc.
4. The Workshop expressed a necessity to further consider the aspects of the possibility to differentiate with accuracy canopy density and height thresholds.
5. The Workshop noted that there is possible confusion between the proposed physiognomic classes for image interpretation and the traditional classes (mainly based on ecological parameters) used by FAO when reviewing existing national forest statistics. The Workshop, therefore, suggested that separate terminology be applied in order to avoid such possible confusion.
6. The Workshop supported the need expressed by FAO that each study carried out using the FAO methodology and classification be correlated by detailed documentation and comments of the interpreter regarding the classes identified and detailed comments on the estimated level of reliability of class delineation.
7. The Workshop concluded that the proposed classification should be adopted, with the considerations expressed above, in this first run of deforestation and forest degradation studies in the framework of FAO Forest Resources Assessment 1990 Project. The experience and results deriving from such studies will allow to further review and improve the classification.

Interpretation Procedure

8. The Workshop recommended that the interpretation of multi-date satellite data be carried out in an interdependent way as proposed by FAO Guideline under Procedure is the most appropriate for a consistent assessment of changes in forest area and quality.
9. The Workshop emphasized the importance of background data such as topographic maps, land use maps, forest type maps, ecological maps, aerial photographs, etc. and field verification for a reliable interpretation of satellite data.

Data Processing

10. The Workshop recommended that the results of the multi-date satellite interpretation be processed in the form of change matrix in order to allow a thorough analysis of the class to class changes in the areas studied.
11. The Workshop considered the data processing procedure proposed by FAO, based on the use of dotgrid and subsequent entry of class codes in computerized spreadsheet files, simple and appropriate to process the interpretation results and produce class-to-class change matrices.

Other Issues

12. The Workshop recommended that such TCDC cooperative action on the issue of forest resources monitoring in the Region be continued with special reference to the countries that could not participate at the present TCDC Workshop, e.g. Pakistan, Sri Lanka, Nepal, Kampuchea, Bhutan, Brunei, Bangladesh, Fiji and other pacific islands countries.
13. The Workshop recommended to establish a regional network between government agencies and other institutions to exchanges and disseminate the forest resources information on a continuing basis.

WORKSHOP ON METHODOLOGY FOR DEFORESTATION AND FOREST DEGRADATION ASSESSMENT IN AFRICA

NAIROBI 25 NOVEMBER – 13 DECEMBER 1991
(first regional workshop)

SUMMARY

Funding Agencies
United States Forest Service (USFS) of The United States Department of Agriculture, Food and Agriculture Organization of the United Nations.

Implementing Agencies
FAO Forest Resources Assessment 1990 Project was responsible for technical implementation and administration; the United Nations Environmental Program (UNEP-GRID) was the host institution.

Participants
Eighteen national resource specialists from 9 countries: Kenya, Mozambique, Nigeria, Sudan, Tanzania, Uganda, Zaire, Zambia, Zimbabwe

Location
UNEP/GRID Headquarters, Nairobi, Kenya.

CONCLUSIONS AND RECOMMENDATIONS

The workshop concluded that the methodology presented by the Project is appropriate and recommended its application for continuous forest monitoring studies.

The following aspects were discussed in detail:

A. Satellite image quality. Background: The images for this workshop were obtained from a variety of sources. The processing was therefore done by several different methods, none of which were documented.

1. Considering the above, the workshop recommended that the data should be produced following a standardized production process, which should be duly documented.
2. The workshop recognized the need for the development of suitable enhancements and band combinations to allow the detection and delineation of woody vs non-woody vegetation types for all ecological conditions.
3. The workshop considered that for the drier areas the standard processing level of satellite images may not meet the minimum acceptable requirements for visual delineation of woody vegetation and recommended that for such areas a more specific approach should be studied and identified.
4. With regard to the drier areas, the workshop recommended that the timing of image acquisition be carefully considered, so that the date of acquisition coincide with the season when the undergrowth vegetation has dried, but the trees have not yet lost their leaves.
5. With reference to the multi-date interdependent interpretation approach the workshop recommended that both images covering the same area be treated in a comparable way to render more consistent the analysis of changes.
6. The workshop recommended that the two images covering a given area be printed at exactly the same scale in order to minimize the difficulties arising from different map projections.
7. The workshop recommended that future assessments of forest change make use of other types of high resolution satellite data, as they become available.

B. Classification

1. The workshop recognized the need for, and recommended the use of, a standard classification of woody vegetation based on physiognomic characteristics in order to allow consistent synthesis of survey results at national, sub-regional, regional and global levels. The workshop noted that it is necessary and appropriate for a country to reserve the freedom and flexibility to further subdivide the standard classes into more specific classes in order to meet the requirements at a national level. Such further subdivision, which must be structured in a consistent hierarchical fashion, could include specific forest formations and floristic groups.
2. The workshop noted that the level of reliability in class delineation strongly depends on ground knowledge and the use of other reference data, such as aerial photographs.
3. The workshop also noted that difficulties in class delineation depended on three main factors:
   • image quality
   • direct knowledge of the area studied
   • level of familiarity with the classification structure and thresholds
   The workshop emphasized that there is need for improvement in all three aspects listed above.
4. The workshop recognized the need to relate the classification to the images, as specifically as possible, and recommended the development of a visual interpretation key for at least the main classes. From the workshop's experience with the variations in image quality, it is recommended that the key be based on texture and patterns, rather than image tone.

C. Interdependent interpretation procedure

1. The workshop recommended the use of an interdependent interpretation approach in order to achieve a consistent assessment of changes in woody vegetation.
2. The workshop noted that the interdependent interpretation approach optimizes the information available in the images, since the interpretation of each image benefits from the information available on the other image.
3. The workshop stressed the importance of having corresponding images produced at the same scale in order to allow a complete and smooth application of the interdependent analysis procedure.

D. Data processing procedure

1. The workshop recommended that the results of the multi-date satellite data interpretation be processed in the form of a change matrix in order to allow a site-specific analysis of the class-to-class changes in the areas studied.
2. The workshop agreed that the processing procedure presented by FAO is simple and appropriate for processing the interpretation results, since it is based on equipment and software which is widely available.
3. The workshop agreed that the processing programmes could be improved or modified as needed in future.

E. Other issues

1. The workshop recommended that a standardized procedure be developed for accuracy assessment of the interpretation.
2. The workshop recommended, in view of point B.2, that a standard ground truthing procedure be developed and that FAO allocate appropriate funding for the task.
3. The workshop recommended that FAO continue to support the member countries in the monitoring of forest resources by:
   1. Continuing training activities and programs.
   2. Networking of regional and national institutions.
   3. Financial and material support such as purchase of satellite images and support of ground truthing activities.
   4. Promoting the implementation of continuous forest cover monitoring at global, regional and sub-regional levels.

WORKSHOP ON METHODOLOGY FOR DEFORESTATION AND FOREST DEGRADATION ASSESSMENT IN LATIN AMERICA

MEXICO CITY 15 MARCH – 2 APRIL 1993

SUMMARY

Funding Agencies
United States Forest Service (USFS) of The United States Department of Agriculture, Food and Agriculture Organization of the United Nations.

Implementing Agencies
FAO Forest Resources Assessment 1990 Project was responsible for technical implementation and administration; the Institute of Geography of the National Autonomous University Of Mexico was the host institution.

Participants
Twenty one national resource specialists from ten countries:
Bolivia, Brazil, Colombia, Equador, Honduras, Mexico, Nicaragua, Paraguay, Peru, Venezuela.

Location
Institute of Geography of the National Autonomous University Of Mexico, Mexico City.

CONCLUSIONS AND RECOMMENDATIONS OF THE WORKSHOP
(English Translation)

1 Regarding the quality of the imagery

The workshop recommends the inclusion of better quality images with less distortion for future assessments, preferably co-registered (historical to recent, or to a common geographical reference)

2 Regarding the classification

• The workshop considers the classification to meet the objectives of forest resources assessment.
• The workshop recommends that the classification be documented by means of detailed studies supported by field verification, aerial photos or imagery of greater resolution.
• The workshop recommends to document the relationship between the general classification (used for the global survey) and the classification systems traditionally used by each country, with the objective of improving of the comprehension and consistency of the general classification.

3 Regarding the interdependent interpretation method

• The workshop recommends the application of this method of interpretation since the interdependent analysis of the two images allows to compensate for poor quality of one of the two images, it minimise the error (due to independent, subjective delineations) and allows a better delineation of the classes.

4 Regarding the processing of the data

• It is concluded that it is a simple technique, which does not require sophisticated equipment and which facilitates distribution of the process at the province and local levels.

5 Practical recommendations for the workshop

• That the spatial analysis be included in the methodology by means of a simple and easily available [GIS] package.
• That the FAO provide tools and common methodologies which will facilitate the analysis and communication between countries.
• The workshop recommends that the FAO orient the activities of the program to central and concrete themes which can be sessions regarding the impacts of deforestation and forest degradation at the regional (continental) level.

6 General recommendations

• The workshop participants recommend that the FAO continue with the forest resources assessment and the distribution of the methodology for multi-temporal monitoring in all countries; and confirm the desire to actively participate in these studies.
• The personal contact established between the participants is considered to be an important result of the workshop; for the exchange of experiences, the application of specific techniques, and everything related to resource assessment.
• It is recommended that FAO actively support this network (of communications) by means of seminars/workshops for evaluation and analysis of the regional results of the FAO forest assessment project, and for the interchange of experiences with the application of the methodology at the national level.

WORKSHOP ON METHODOLOGY FOR DEFORESTATION AND FOREST DEGRADATION ASSESSMENT IN AFRICA

YAOUNDE 28 FEBRUARY – 18 MARCH 1994
(second regional workshop)

SUMMARY

Funding Agencies
The European Community (EEC) and the Food and Agriculture Organization of the United Nations (FAO).

Implementing Agencies
FAO Forest Resources Assessment 1990 Project was responsible for technical implementation and administration; the National Office of Forest Development (ONADEF) was the host institution.

Participants
Twenty seven National resource specialists from twelve Central and West African countries:
Angola, Cameroon, Central African Republic, Congo, Equatorial Guinea, Gabon, Guinée, Ivory Coast, Liberia, Nigeria, Sierra Leone, and Zaire.

Location
Ecole Nationale Superieure Polytechnique, Yaoundé, Cameroon.

CONCLUSIONS AND RECOMMENDATIONS

CONCLUSIONS

• Concerning the quality of the satellite images, the participants acknowledge FAO's efforts in providing good quality images of the region.
• The procedure of interdependent interpretation, which allows to work on multidate images from one location was judged to be the most reliable methodology on account of it's coherence and logic.
• The data compilation procedure was found satisfactory and the application of Lotus Software, compatible with other common software was found appropriate. It also enables the presentation of results in the form of change matrices which are easy to produce and interpret.
• The FAO vegetation classification caused some difficulties concerning its terminology compared to the various National classifications.
• Considering that the FAO methodology allows vegetation state and change assessment, the participants recognize the need and opportunity to install in each country remote sensing facilities.

RECOMMENDATIONS
Following the above conclusions, the participants have adopted the following recommendations:

• FAO should make efforts to provide satellite images to the Project countries.
• FAO should continue to disseminate the methodology for deforestation and forest degradation assessment.
• Regarding the introduction and standardization of the FAO classification for the pantropical zone, the workshop recommends to every country of the region to correlate their National classifications with the FAO classification.
• As a follow up to the workshop, the participants recommend that each country should make an effort to install computer equipment. In particular, they recommend FAO to encourage these National initiatives and provide assistance.
• For a better transfer of the methodology the workshop recommends the creation of permanent regional training centres.