Previous Page Table of Contents Next Page



A goal of any forest resource monitoring program must be to address change in biomass as well as its state. Estimating national emissions of CO2 on a regular basis or monitoring CO2 mitigation projects also require information on change in biomass, both losses or gains. Because no real attempts have been made in the past to estimate the state of biomass in most tropical countries, estimating its change becomes a very difficult task. To substantially reduce errors in estimates of the state of biomass and its change, a new approach is needed that will incorporate in its design methodology for monitoring biomass density and its change more directly.

One way to estimate changes in biomass density and total biomass is to combine data on forest area change, by vegetation type, with biomass density values for the same vegetation types. Vegetation-cover transition matrices produced from the interpretation of remote sensing imagery, such as were produced by the FRA 1990 Project (FAO 1995) can provide data on forest-area change. This approach will only provide an "educated estimate" of biomass change because although the area change will have a high degree of statistical reliability at the regional and global levels, there will have been little to no improvement in the forest biomass density values.


A more comprehensive and reliable approach for estimating biomass change is to combine new field studies with analysis of high resolution remote sensing imagery. The remote sensing efforts would be used to delineate forests into various distinct biomass strata. Then, using a statistical design, permanent plots in these forest strata could be established, for example, by using the methods described in Lund (1992). For estimating biomass density directly, stand tables are sufficient with use of the generic biomass regression equations given in Section 3.2, or locally derived ones where resources allow (see Section 4.). At least two measurements on permanent plots are needed to estimate biomass change. These measurements should preferably be a minimum of 5 year apart, particularly in forests vulnerable to change. However, results from a one-time survey combined with change in area data could provide a more reliable pan-tropical estimate of biomass change than any one available at present.

The established permanent plots must be remeasured regularly and be integrated into a framework of continuous forest monitoring. This approach will provide the most accurate and precise estimates of total biomass and biomass density change for policy makers.


A key component of any resource monitoring program is to try and understand the processes affecting change. A better understanding of the processes involved will lead to improved planning and control of forest biomass degradation and ultimately economic savings. Furthermore, understanding the mechanisms involved in forest biomass change will enable predictions about future trends in forest resources to be made which will be directly applicable to the local demands, timber trade, and global issues of concern such as global change and biodiversity.

The basic approach would be to develop a model that relates loss in biomass density, e.g., the degradation ratios, to readily measurable indices of human activity and bio-physical parameters. A similar model could be developed for forest area loss, so that a combination of the two models would give total biomass change. The availability of modem technology of remote sensing and GIS makes this task feasible. The models should include the spatial component of the landscape and human activities.

The approach would be to use the GIS data layers such as those identified above in Section 6.1 and new ones as they become available. New data layers will likely result from remote sensing and field studies. These new GIS data layers could be combined in models to obtain various indices that most likely influence or are influenced by human use of the land such as: (1) bio-physical and climatic factors, (2) fragmentation indices (e.g., perimeter/area ratios of forest area), (3) transportation networks (e.g., roads, rivers, railways), (4) population density and it rate of growth, and (5) socio-economic and political factors (e.g., land tenure, market availability, GNP/capita). How bio-physical and climatic factors regulate potential forest biomass is reasonably well understood. What is not well understood is what measure of human activity should be used in the models. To date, percent forest cover and degradation of forest biomass density have been shown to be highly correlated with population density at subnational levels when stratified into ecological zones. But, as stated above, the high correlation does not imply causal relationships. We know that deforestation and degradation cannot continue indefinitely; eventually a limit is reached, beyond which forest cover and biomass density levels. Moreover, other socioeconomic factors will eventually become active and reverse the trends as has been the case in industrialized countries.

To bring about these new directions in estimating state and change in biomass will require a concerted effort and expenditure of resources. A first step is to increase training in GIS analysis, biomass estimation, field-inventory operations, and remote sensing analysis, in all countries. This could be accomplished by developing networks between forestry personnel in tropical countries and institutions performing these kinds of activities. A sharing of data bases among countries is needed also so that the issues of forest cover and biomass change can continue to be monitored with improved understanding of the mechanisms and processes that influence the changes.

Previous Page Top of Page Next Page