The regions of the independent remote sensing aided survey could be groups of countries, continents or sub-continents. For optimal sampling strategies, the regions can be further stratified into sub-regions on the basis of the vegetation zones, information needs, existing information and the possibilities of the control inventory, as well as remote sensing data (see, FAO 2001 b and Appendix 1). Regional designs are relevant because of specific information needs, technical reasons and variability of the target variables. Aggregation to global level is, however, the goal of the RSFS survey.
Analysis units can be determined on the basis of the variability of the variables of interest, e.g., percent forest or forest area change. The information from earlier assessments can be utilised in identifying the regions and analysis units. The land areas and forest areas of continents are given in Table 1(FAO 2001 a).
Table 1: Forest area by region 2000 (FAO 2001 a)
Region |
Land area |
Forest area |
% of land area |
% of all forests |
Net change 1990-2000 |
mill. ha |
mill. ha |
mill. ha/year | |||
Africa |
2978 |
650 |
22 |
17 |
-5.3 |
Asia |
3085 |
548 |
18 |
14 |
-0.4 |
Europe |
2260 |
1039 |
46 |
27 |
0.9 |
North and Central |
2137 |
549 |
26 |
14 |
-0.6 |
America |
|||||
Oceania |
849 |
198 |
23 |
5 |
-0.4 |
South America |
1755 |
886 |
51 |
23 |
-3.7 |
Total |
13064 |
3869 |
30 |
100 |
-9.4 |
Africa is divided into 6 sub-regions, Asia into 5, Europe into 4, North and Central America into 3, Oceania into 2 and South America into 2 sub-regions in FAO FRA 2000 main report (FAO 2001 a). These sub-regions, possible combined with vegetation zones, could be suitable sampling design units for RSFS. The optimisation with respect to changes may require more detailed stratification of the target areas. The changes are often clustered and local variation in changes is high (Drigo 2002). The highest relative change rates of the regions of Table 1 occur in Africa, South America and Oceania. A challenging task is to get a priori information about change hot spot areas. Remote sensing data, e.g., medium resolution data (Chapter 3), can be utilised in this task.