2.1 Organization in relation to NFA design

In designing NFAs there are many optional ways to increase the theoretical efficiency of the inventory by incorporating sophisticated statistical techniques (e.g. Thompson 1992, Schreuder et al. 1993). For more details see Sampling design. However, there is clearly a trade-off between theoretical efficiency and simplicity of the design. The latter is very important from the organizational point of view of the NFA. All the following issues advocate the choice of a simple, robust and straightforward design:
  1. The availability of skilled personnel within narrow subject fields generally is limited, and the more complex designs are chosen, the more vulnerable the organization will become with regard to dependency on key personnel.
  2. A simple design implicitly gives more people intellectual access to data, and it will be possible for many people to perform analyses based on the NFA data, which is important.
  3. With a simple design, most stakeholders will understand how forest and land-use statistics are derived and thus it is likely that there will be a better basis for common agreement on the results obtained.
  4. Complicated designs, possibly adopted to reduce costs in fieldwork, may become very expensive, since complex analysis tools need to be developed to analyze the data. Also, there is an increased risk that the results will be erroneous due to misunderstandings.
  5. With complex designs, more money has to be spent on competence development.
To sum up, in many cases a simple and straightforward design approach is to be favoured to a complex but theoretically efficient design. Thus, the organization of the NFA and the competence level of people likely to work with the NFA need to be considered during the design phase.

A closely related issue concerns the amount of variables included in the survey. Many of the five issues above are valid also in this regard, e.g., the need for training and competence development will be higher when more variables are introduced in the inventory. Also, there may be a risk that focus is lost from the most important variables, and that data quality will be reduced.

Another important design issue concerns whether or not NFAs should be carried out continuously (i.e. every year), averaging data over several years when estimating the current state of some feature, or if the NFA should be carried out only at certain intervals in time. From the organizational point of view, many arguments can be found that favour a continuous approach:
  1. With a continuous inventory, it is possible to hire personnel on a long-term basis. Thus, it is likely that the organization may be able to attract more skilled personnel than would otherwise be the case.
  2. The time allocated to training is likely to decrease considerably. With inventories run only at certain intervals, many people are likely to be new in the organization each time the inventory is run.
  3. It will be possible to maintain continuity in the inventory with regard to methods and assessment techniques used. This reduces the risk of obtaining varying levels of surveyor induced systematic errors in the NFA results between different time points.
On the other hand, there may be drawbacks in running a continuous inventory as well. One example is the increased difficulty of implementing changes in the NFA. Of course, there is also an aspect of costs and priorities. In many developing countries, for example, the issue is simply a matter of doing the best possible work within given project resources and time (Velle, 1995).

In case an NFA organization has to be established to run an inventory on a non-continuous basis, i.e. that the organization would need to down-size between the times the NFA actually is carried out, it may still be possible to establish continuity in the organization by linking the NFA organization closely to a more permanent body, e.g. a university or a forest research institute. In such a case, one could link the NFA activities with other periodic activities, implying that many of the positive features of running an inventory on a continuous basis can still be achieved. Another possibility is that the core part of the organization is stable, concentrating on data acquisition during some periods and data analyses during other periods. In this case, only the field staff would need to be employed on a discontinuous basis.