4. Conclusions and recommendations

During the review, old sources used to develop previous estimates were thoroughly analyzed and data were recalculated in more accurate ways. At the same time, for certain countries (Namibia for example) new sources of information were located and used in order to improve the estimates. As a result a certain discrepancy with the old FRA 2000 estimates was found. The difference has to be considered the consequence of the constant effort of FRA to improve the assessment process and consequently the reliability and consistency of estimates provided.

Excluding 54 industrialized countries (TBFRA 2000 regions) which were assessed by the FAO-UNECE, a total of 159 developing countries were analyzed and reviewed.
It is worth to mention that the level of availability and reliability of information is still very low in the developing countries and the reviewed volume and biomass estimates still embody a certain degree of uncertainty.

As already explained, estimations on wood volume and woody biomass were based on various types of information sources, which reliability and precision affected somehow the consistency of the results.

When available, data resulting from partial or national forest inventories were highly variable in terms of quality, spatial, thematic and temporal resolutions. Problems of comparability of national data and reliability of aggregated results arose mainly because of differences in the national systems of nomenclature. The differences in vegetation classification systems were eliminated by harmonizing definitions with the FRA 2000 forest related definitions while differences in measurement techniques and definitions of volume and biomass were harmonized using standardized parameters (VOB 10, total aboveground biomass, etc). Differences of mean volume and mean biomass per hectare between the reference periods were assumed to be negligible.
Existing data coming from partial or national forest inventory were found for 49 countries (23% of the total number of countries): in general data coming from forest inventories although presenting sometimes, problems of interpretation of the data, were the most reliable information source.

Data coming from biomass assessment studies carried out for different purposes (carbon sequestration, fuel wood resources etc) were used to develop estimates for 11 countries (5.16% of the total number of countries): depending on the methodology used to estimates the woody biomass (remote sensing, field measurements) data showed an intermediate level of reliability, even if assumptions were often necessary.

General forest assessment studies (e.g. statistical yearbook, forestry papers) were used as source of information for 30 countries (14% of the total number of countries): due to the fact that provided data were often not referenced and no explanations were given on the type of volume provided (total volume, stem volume, including or excluding bark, minimum DBH used) results showed a higher level of uncertainty.

No reliable information was available for 28 countries (13.15% of the total number of countries) and an expert opinion had to be used in order to develop estimates on wood volume and woody biomass.

Finally no information sources were available for 41 countries (19.25% of the total number of countries). It is worth to mention that, as showed in Table 6, these countries bear only 0.19% of the total world forest area and their lack of information do not affect that much the reliability of the results, from a global prospect.

Table 6: Type of information source and representatively.

In general for all those countries with lack of complete or reliable data, undertaking of the national forest resources assessment is strongly recommended.

The problems related to information scarcity and low reliability emphasize the importance of continuing to work at national level to support developing countries to improve their information on the forestry resources

Support to National Forest Assessments activity being developed as part of the Forest Resources Assessment Programme may help in generating reliable information on a wide range of forest and tree parameters on the basis of a harmonised set of variables and a vegetation classification system and standard forest and tree survey sampling design for continuous monitoring.

Launch a study that can clarify vegetation type/species specific estimates of VEF, BEF and WD at country level, including factors for capturing below ground biomass and C content may be also recommended.

Since strict referencing of data source is essential for the development of the assessment process, volume and biomass data will be included in the Forest Information System (FORIS) as soon as the system itself will be ready for this type of data entry; states for volume and biomass will be created for each country as it has been done for the forest area parameter.