5.1.1 The NPP demonstration project was initially proposed at the meeting of Experts of Ecological Networks in Guernica, Spain, 1997. The NPP project has two primary goals: first to distribute a global standard NPP product to regional networks for evaluation; and to translate this standard product into regionally specific crop, range and forest yield maps for land management applications. The project will demonstrate the potential of a limited network of surface sites to generate a reliable, useful product with global coverage and local relevance, rapidly and efficiently, by adopting the GHOST hierarchical sampling approach and by using models to combine in situ and remotely sensed data. The project is designed so that any network or site, regardless of its level of sophistication, can make a useful contribution to the project. The goal is to complete a demonstration project that would be of value to the participating GT-Net regional or national networks/sites, and to allow the development of the process for contributing site data to a central archive and utilising site capabilities to validate satellite imagery. The primary source of satellite data will be the AM-1 satellite of the Earth Observing System (EOS), to be launched in 1999 by the NASA.
5.1.2 It is envisaged that site data would be provided to the U.S. LTER Network Office for incorporation into the TEMS database and land cover and Leaf Area Index (LAI) validation data would be provided to the U.S. LTER Network Office for archival. Satellite imagery of land cover and LAI would be provided to sites by the U.S. LTER Network Office. In addition, the project will use the FLUXNET global network of eddy covariance flux towers to translate the EOS NPP product into net ecosystem productivity (NEP), or total net ecosystem CO2 flux, for use in global climate and carbon cycle models. In this context, the NPP project will inaugurate a globally distributed tracking of NPP and NEP which is critical to global change monitoring. As GT-Net achieves global coverage, it will quantify the significant trends in biospheric productivity, including NPP responses to desertification and land degradation trends, interannual climate variability, CO2 fertilisation, pollution effects, etc. This plan also incorporates the principles of GHOST, the global hierarchical observing strategy, in that the regional NPP drivers will be done as a tier 4 activity, regional NPP validation a tier 3 activity, and the FLUXNET towers function as tier 2 intensive continuous NEP study sites. Finally, the plan offers two end products, one of high practical utility at the local level, the other of high significance at the scientific level.
5.1.3 The project will extract the global 8-day NPP product from the U.S. NASA EOS Data Information System for dissemination to participating surface networks. The U.S. LTER Network Office will lead this effort, including archiving site data and extracting appropriate satellite coverage for individual sites. Specifically, EOS is computing a NPP product at 1 km for the entire global vegetated land surface every 8 days from the MODIS sensor and ancillary data. The initial EOS NPP algorithm is based on the relationship between time integrated absorbed PAR and NPP.
5.1.4 Each participating site should aim to collect the following input variables. Failure to collect the full set does not render the site useless, since the data, that are available, can be used for independent validation of the EOS-derived global data sets. The data must be collected to be representative of a minimum area of 3 km x 3 km. This is because the calculated NPP product will initially have a resolution of 1 km by 1 km, with a positional error of up to 1 km. Future products will have a resolution of 250 m x 250 m (see Annex X for the implementation steps of the NPP project).
Recommendation 6: TOPC recommends that the NPP project should be executed in accordance with the above plans. TOPC should maintain contacts with the projects as it progresses, and should use the lessons learned to refine plans for other observing networks and for the compilation of global data sets.
5.2.1 Dr. Cihlar presented information on the Global Observations of Forest Cover (GOFC) project on behalf of the project co-ordinator, Dr. Frank Ahern from the Canada Centre for Remote Sensing (CCRS). The presentation was based on the current planning document and an update regarding recent activities and plans.
5.2.2 The GOFC project was proposed as part of several initiatives of IGOS in February, 1997. The objectives and an initial plan were defined at a July, 1997 workshop in Ottawa, Canada. As a contribution to the implementation of an ongoing global forest monitoring system, GOFC will:
Produce high quality, multi-resolution, multi-temporal global data sets and derived products of forest cover and attributes;
with particular attention to areas of rapid change and fragmentation;
to be repeated for quantitative analysis of variation on a three to ten year cycle;
with associated regional applications and methodological investigations;
for the benefit of multiple user communities;
and ultimate transition to routine operational use.
5.2.3 GOFC plans to generate a series of products for diverse user communities, including: international environmental conventions and agreements, international policy and operational institutions, national and sub-national institutions, and the scientific community. The products are aimed to serve various applications: global and climate change, sustainable forest management and multiple land use, biodiversity, and others. The principal products to be generated are land cover and land cover change (at detailed spatial resolution), land cover at medium resolution for modelling purposes, net primary productivity, biomass, fire scars, harvest, leaf area index, fraction of photosynthetically active radiation, and the total amount of photosynthetically active radiation. GOFC is to take place over the next five years, and is expected to form the basis of an ongoing satellite-based monitoring system for the global forest. The project planning is led by the CCRS with the support of the Canadian Space Agency. Numerous space agencies have agreed to participate by contributing data or other resources.
5.2.4 In the subsequent discussion TOPC addressed the issues of product characteristics, product generation and distribution plans, and linkages to existing programs.
5.2.5 Dr. Townshend relayed results of the discussion at the last Strategic Implementation Team (SIT) meeting which considered it essential that GOFC be embedded in existing international programs. In particular, SIT requested that GOFC meet the needs of GCOS and GTOS as established by TOPC. To this end, GOFC should report to GCOS and GTOS through TOPC. The project should also meet the needs of IGBP as defined by IGBP DIS.
Recommendation 7: The organisational structure of GOFC should be modified to include formal TOPC representation in GOFC planning and implementation. TOPC representative should represent the authoritative requirements of GCOS/GTOS climate-related applications, and should report to TOPC on GOFC developments.
5.2.6 The difficulty of defining acceptable classification systems was discussed. A key requirement here is that the classification legends must be designed to meet the needs of specific user groups. Since this is difficult, and given that any fixed limits used render the products of limited usefulness, it is recommended that the GOFC should aim to define products quantifying continuous variables. These should include canopy cover, leaf area duration, and canopy height. It was pointed out that by avoiding the need for identifying arbitrary class limits, such variables permit various users to employ the same product for their specific purposes.
Recommendation 8: GOFC should consider inclusion of continuous variable products, especially canopy cover, leaf area duration, and tree height. GOFC should also actively promote and facilitate further research to generate such products, where appropriate.
5.2.7 To make the products widely useful in the tropical areas, it is important that the information extracted from satellite data allow determination of the extent of tree crops. Similarly, GOFC planning and product generation should include expertise with data analysis from sparse forest canopies, especially woody savannahs. It was suggested that GOFC should make direct contact with the Centre for International Forestry in Indonesia which is concerned with forestry issues at the international level and is particularly concerned with sustainable development. Further, it is essential to ensure that the product formats do not hamper the accessibility of the GOFC products in various countries of the world.
Recommendation 9: GOFC should ensure that the planned products meet the needs of the principal user groups. The principal user groups should be identified for the various planned products. Further, hardcopy as well as digital products should be planned.
