1. The North American Forest Commission (NAFC) provides a framework for Canada, Mexico and the United States to address issues of common interest in the forest sector. Several recent developments led the NAFC to consider improved collaboration in forest inventory, assessment and monitoring, including:

· The NAFC countries face similar challenges in implementing the Montreal Process criteria and indicators for sustainable forest management.

· The NAFC countries are each contributing information to the global forest resources assessment (FRA 2000). There may be value in a common approach to implement a remote sensing survey to monitor forest change processes in North America.

· Each NAFC country is trying to improve its forest inventory and monitoring systems to provide more complete and timely information about forest resources and trends. There may be an opportunity for regional collaboration.

2. In March 1998, the NAFC Bureau of Alternates decided to sponsor a workshop on forest inventory, assessment, and monitoring to address these issues. The United States offered to host the workshop and to report the outcomes at the November, 1998 meeting of the NAFC in Villahermosa, Mexico. The workshop was held at the USDA Forest Service Remote Sensing Applications Center in Salt Lake City, Utah, July 7 - 10, 1998.

WORKSHOP OBJECTIVES

· Develop a common understanding of existing forest resource assessment programs in the three NAFC countries: Canada, Mexico, and the United States.

· Develop an action plan for the North American remote sensing component of FRA 2000, including agreement on who will do the work, how it will be funded, and a timetable for completion.

· Identify opportunities for increased collaboration among NAFC countries in forest inventory and monitoring systems. Develop an action plan for future collaboration, including mechanisms for coordination.

· Consider Mexico's special case, including a proposed Mexico deforestation study. Develop an action plan as appropriate.

REVIEW OF FOREST RESOURCE ASSESSMENT PROGRAMS IN NORTH AMERICA

3. The first part of the workshop featured presentations on forest inventory, monitoring and assessment programs in North America, in order to bring participants to a common level of understanding of the existing programs and capabilities of the three countries and key institutions. The presentations are summarized in Annex 2.

· Pauline Myre - Forest assessment programs in Canada

· Bernard Herrera - Forest assessment programs in Mexico

· Andy Gillespie - Forest assessment programs in the US

· Ashbindu Singh  - UNEP GRID assessment programs

· Zhi-Liang Zhu  - USGAS EROS Data Center forest & ecoregion mapping

· Aurelio Fierros  -  Montreal Process criteria and indicators

· Frank Ahern  - Global Observation of Forest Cover

· Jim Strittholt  - WWF study of forest fragmentation in North America

· Celedonio Aguirre-Bravo  - Ejido El Largo project & North American Symposium on Inventory & Monitoring

· Robert Davis - FAO, FRA 2000

· Ray Czaplewski -  Sampling design for FRA 2000 remote sensing survey

· Jose Palacio - Analysis of land use changes in Mexico

· Jean Mas -  Analysis of forest change in Campeche, Mexico

FRA 2000 REMOTE SENSING SURVEY FOR NORTH AMERICA

4. Context: The governments of Canada, Mexico and the United States, along with about 80 other countries, endorsed FRA 2000 at the FAO Committee on Forestry (COFO) in March, 1997, and at the Intergovernmental Panel on Forests (IPF). The IPF proposals for action were approved by the UN General Assembly Special Session (Rio +5) in June, 1997. In addition, Canada and the United States agreed to support FRA 2000 in the G-8 Action Plan for Forests, approved at the G-8 summit in May, 1998. Implicit in the endorsement of FRA 2000 is a national commitment to provide information about one's country for inclusion in regional and global studies and reports.

5. Both COFO and IPF requested that a globally consistent methodology be used in FRA 2000 so that information for developing and developed countries would be comparable. As in 1990, some of the information will be provided by countries based on existing national forest inventories, but information from existing inventories is inconsistent due to the wide range of national forest inventory approaches and capabilities. FAO was mandated to undertake a global remote sensing survey using standardized methodologies.

6. Remote Sensing Survey methodology: Forest areas will be identified in each country using AVHRR imagery. Roughly 10% of the forest areas will be selected for more intensive analysis using high resolution imagery such as the Landsat 5 Thematic Mapper (20% in high risk tropical areas). Imagery for 1980, 1990, and 2000 (or as close to those dates as possible) will be compared using a standard hard copy-based method, which can be done consistently in all parts of the world. Changes in forest cover will be mapped and quantified for each sample scene. The statistical sample is sufficient to draw conclusions about change in forest cover at the continental and global level. For large countries such as the US, Canada, and Russia, the sample will be large enough to draw conclusions at the country level. However, FAO will only use the statistics for reporting at regional and global levels.

POTENTIAL BENEFITS FROM A NORTH AMERICAN REMOTE SENSING SURVEY:

7. Workshop participants identified the following potential benefits:

Global

· By definition, "global" conclusions from FRA 2000 must include all countries.

· A major issue at IPF was the need for consistent global information about forest changes and indicators of sustainable management; it was agreed that a global remote sensing survey would be a key step toward achieving this consistency, especially given the vast differences in information reported by countries from existing sources.

· The credibility of conclusions about a given country reported by FRA 2000 depend on the use of globally consistent methodologies, especially for remote sensing.

Regional

· The use of consistent methodologies for collecting forest resource information will enable results to be compared at the regional level, partially overcoming differences due to different inventory approaches used by the three North American countries.

· The FRA 2000 remote sensing survey could be a first step toward broader collaboration in remote sensing and assessments, with potential benefits from economies of scale. An example is a consistent regional sampling grid, which could be used for many other scientific purposes.

· All three NAFC countries plan to increasingly use remote sensing and GIS technology in assessments at all levels. The experience gained from the first national remote sensing surveys would provide R&D benefits which are difficult to anticipate.

· Technology transfer between the NAFC countries and capacity building will be inevitable by-products of a regional approach. Each country stands to benefit from the lessons learned by the others.

· The NAFC countries could elect to prepare a regional report addressing issues of regional importance that are not included in the global report.

Country

· By participating in the FRA remote sensing survey, a country contributes to a process which provides consistent data for all countries, which are in turn in demand by the citizens of the country. The FRA assessment is the only global process which accomplishes this. In a sense, a country is exchanging its knowledge about its own forest resources with other countries, to the benefit of the citizens of all countries.

· Although FAO does not plan to draw national-level conclusions based solely on the remote sensing survey, the U.S. and Canada are large enough that the FRA 2000 remote sensing methodology will enable the countries to draw statistically valid conclusions about forest change; in the case of Mexico, the sample will need to be intensified in order to allow the country to draw national-level conclusions.

· The remote sensing survey could provide validation of national estimates of key parameters derived from other sources. An example is biomass and estimates of carbon storage and carbon cycling, which are key components of the Kyoto protocol.

Information to be Generated by the Remote Sensing Survey

8. The remote sensing survey is the only component of FRA 2000 for which all countries will use comparable data and methodologies; it will be the key to assessing forest change processes. Specific outputs include:

· Change in forest cover over the period 1980 - 1990 - 2000

· Biomass fluxes associated with changes in land cover for the same period

· Forest frangmentation and related multi-temporal and spatial analyses

· Countries will participate in the analysis of their own remote sensing data; it is anticipated that additional uses for the information will result as national capacities are increased.

INVENTORY AND MONITORING

9. Each NAFC country has an existing forest inventory and monitoring system, along with a considerable number of people skilled in inventory techniques and the use of new technologies, including remote sensing and GIS.

10. However, existing inventory and monitoring are not satisfactory in any of the NAFC countries in terms of meeting the demand for more complete and timely information about a wide variety of indicators of sustainable forest management. In the course of the workshop, it became clear that each country faces similar challenges, even though there are considerable differences in their existing systems.

11. Summaries of the three country presentations, in brief:

· Canada's current National Forest Inventory is a periodic compilation of existing inventory material from across the country. Basically, stand level data are converted/ harmonized to a national classification scheme, and aggregated to the mapsheet, provincial and national levels for storage, analysis and reporting. Canada is considering a new National Forest Inventory design consisting of a plot-based system of permanent observational units located on a national `grid.' The objective of the new inventory would be to assess and monitor the extent, state and sustainable development of Canada's forests in a timely and accurate manner. In particular, it should provide data for the criteria and indicator processes to monitor sustainable development, and data for national and international inquiries.

· Mexico has undertaken several national inventories, most recently in 1994, each time using different methodologies. As a result, it is difficult to compare the results of the different inventories to monitor trends over time. Mexico is also considering establishing a national system of forest inventory plots.

· The US has a national system of forest inventory plots (FIA) and a relatively new system of forest health monitoring plots (FHM), which addresses a larger number of parameters than does FIA. However, plots are visited only on a ten-year cycle, and there are holes in the system - for example, information from plots on national forest lands has not been consistently gathered. The US is considering steps to combine FIA and FHM, and to increase the periodicity of systematic sampling.

MEXICO DEFORESTATION STUDY

12. Prior to the workshop, Mexico asked the US Forest Service for assistance with a study of deforestation. The last day of the workshop was set aside for discussion of this issue.

13. Several previous studies, including FRA 1990 and the 1994 national forest inventory, indicated that deforestation is taking place at alarming rates in Mexico, with the highest rates in the tropical forests in the South. A recent nation-wide study of changes in land uses using Landsat MSS data, carried out by UNAM with support from the US Environmental Protection Agency (EPA), has been helpful in identifying trends in land uses at the national level. But the study was not successful in providing accurate estimates of deforestation rates in different areas, nor in quantifying the causes of deforestation or forest degradation. In particular, none of the previous national studies have produced conclusive results about large areas of land which are characterized by forest fragmentation and the gradual conversion of forests to other land uses.

14. It is not known the extent to which deforestation rates have been affected by changes in government policies in the 1990's which were specifically intended to slow the rate at which forests were being converted to agricultural uses, especially in the South. Mexico's decision makers need better information about the location and causes of deforestation in order to decide if policies need to be changed and if enforcement needs to be strengthened.

15. In the ideal world of unlimited resources, Mexico would conduct a new wall-to-wall study of deforestation in all forested areas of the country. However, such a study is currently beyond the financial capabilities of the Mexican government or the US Forest Service. As an alternative, it was agreed to try to "kill two birds with one stone" by using the FRA 2000 remote sensing survey to analyze changes in forest cover between 1980 - 1990 - 2000 in selected areas, and to intensify the number of samples (more than 10%) by selecting additional scenes in areas where deforestation is known to be a serious problem as a result of previous studies ("hot spots").

16. It was further agreed that there is a need for improved collaboration among SEMARNAP, UNAM, INIFAP, and other key organizations in Mexico, including a network of universities and institutions engaged in research using remote sensing and GIS technology. There is a need to avoid duplication and to optimize the use of limited resources to assess forest resources and analyze deforestation. Aurelio Fierros, SEMARNAP, will speak with Victor Sosa about convening a workshop in Mexico to discuss this possibility. The US Forest Service and FAO will be invited to participate.

17. Tom Bobbe, Coordinator of the USDA Forest Service Remote Sensing Applications Center, offered to assist Mexico to obtain updated satellite imagery required to carry out the proposed deforestation studies.

CONCLUSIONS AND RECOMMENDATIONS

18. FRA 2000 Remote Sensing Survey

· The United States and Mexico will implement the remote sensing survey for FRA 2000 in their respective countries, subject to confirmation of funding. Canada will make a decision after further discussions at home. In the US, the Forest Inventory and Analysis (FIA) program will take the lead, with technical assistance from the Remote Sensing Applications Center (RSAC). In Mexico, the National Forest Inventory unit of SEMARNAP will take the lead, with technical assistance from the Autonomous University of Mexico (UNAM). The US Forest Service will assist Mexico in acquiring satellite imagery.

· FAO will review the proposed land cover classification scheme and develop a common approach for use by both developing/tropical countries and developed / temperate and boreal countries. Several workshop participants suggested that FRA 2000 might use a two-step process: first, classify each area of land according to land cover; second, describe and quantify changes in land uses. An ad hoc NAFC committee will assist FAO with this effort: Jose Palacio (Mexico), Mike Hoppus (US), and Mark Gillis (Canada).

· A second ad hoc NAFC committee will advise FAO on several unresolved issues related to FRA 2000 remote sensing methodology, including a sampling grid, interpretation (manual vs. computer-assisted), sample unit size, and sampling intensity. Representatives from the EU Joint Research Centre and other institutions will also be consulted. NAFC consultants will include: Ray Czaplewski, Tony Olsen, and Andy Gillespie (US); Jean Francois Mas (Mexico); and Mark Gillis (Canada).

· It was agreed that there is a need for a consistent approach for classifying areas which are partially forested or "fragmented forest." Jose Luis Palacio, Mexico, volunteered to assist FAO by pilot-testing modifications to the forest classification scheme and developing a classification manual for use by interpreters. This work will be coordinated by Robert Davis and Rudi Drigo, FAO.

Inventory and Monitoring

· Representatives from the three NAFC countries agreed in principle to develop a common approach for locating forest inventory plots, and to work toward greater consistency and collaboration on forest inventories and monitoring. Initially, Andy Gillespie (US), Mark Gillis (Canada), and Bernard Herrera (Mexico) agreed to work together on an ad hoc basis until such time that a more formal mechanism is established.

Mexico Deforestation Study

· The FRA 2000 remote sensing survey will form the framework for a new deforestation study in Mexico. Forest cover in selected areas will be compared using high resolution satellite images for 1980, 1990 and 2000 (or surrogate dates). The 10% sample for the North American survey will be extended in Mexico to include known "hot spots" for a more intensive study of land use changes and the causes of deforestation.

· Representatives from SEMARNAP offered to convene a consultative process within Mexico to bring together key people in universities, institutes, and national and local forest management agencies with the goal of improved collaboration on inventory and monitoring. The FRA 2000 remote sensing survey / Mexico deforestation study will provide an excellent opportunity for synergy among key organizations.

Recommendations to NAFC

· The report of the workshop will be submitted to the North American Forest Commission for consideration at the 19^th session in Villahermosa, Mexico in November, 1998.

· The NAFC is requested to establish a study group on forest inventory and monitoring to provide a formal mechanism for collaboration at the regional level. (The workshop consensus was that "forest inventory and monitoring" was prefered over "forest resource assessments" as a proposed name for the study group.)

Workshop participants noted that inventory and monitoring provide the information to conduct research and to make policy and management decisions based on science. In particular, information about large forest areas is critical to monitoring changes in the health of forest ecosystems and for tracking criteria indicators of sustainable forest management. Many issues relating to forest sustainability transcend national boundaries. NAFC study groups provide a logical mechanism for collaboration, since a structure for management and policy oversight is already in place.

Tasks for the study group might include:

· Analyze opportunities for increased consistency among the strategic inventory and monitoring systems in the NAFC countries.

· Develop and implement collaborative programs and projects.

· Work together to complete FRA 2000 and future global assessments.

· Develop linkages with other NAFC study groups.

· Identify current and future trans-boundary information needs.

LIST OF PARTICIPANTS - NAFC FOREST ASSESSMENT WORKSHOP

Canada

Frank Ahern, Canada Centre for Remote Sensing, Natural Resources Canada

Mark Gillis, Manager, National Forest Inventory, Canadian Forest Service, Natural Resources Canada

Brian Haddon, Manager, National Forestry Database Program, Canadian Forest Service, Natural Resources Canada

Pauline Myre, Director, Policy and Planning, International Affairs, Canadian Forest Service, Natural Resources Canada

Bruce Pendrel, Acting Director, Forest Resources, Canadian Forest Service, Natural Resources Canada

Mexico

Aurelio Fierros, Director of Commercial Plantations, SEMARNAP

Jean Francois Mas, GIS Specialist, Centro EPOMEX, University of Campeche

Jose Luis Palacio, Director, Instituto del Geografia, Autonomous University of Mexico

Bernard Herrera y Herrera, Sub-Director, National Forest Inventory, SEMARNAP

USA

Celedonio Aguirre-Bravo, Mexico Research Coordinator, USDA Forest Service

Tom Bobbe, Manager, Remote Sensing Applications Center, USDA Forest Service

Ray Czaplewski, Project Leader, Multi-Resource Inventories, USDA Forest Service

Andy Gillespie, National Forest Inventory Program Leader, USDA Forest Service

Mike Hoppus, Remote Sensing Applications Center, USDA Forest Service

Doug Kneeland, Office of International Programs, USDA Forest Service

Tony Olsen, US Environmental Protection Agency

Doug Powell, National Monitoring and Evaluation Coordinator, USDA Forest Service

Mike Sieg, Deputy Director, Inventory & Monitoring Institute, USDA Forest Service

Brad Smith, Forest Inventory and Analysis Program, USDA Forest Service

Jim Strittholt, Conservation Biology Institute

Zhi-Liang Zhu , USGS EROS Data Center

International Organizations

Robert Davis, Forest Resources Assessment Program Coordinator, FAO

Ashbindu Singh, UNEP Environmental Information & Assessment Program