Evaluating scale dependence of ecosystem service valuation: a comparison of NOAA-AVHRR and Landsat TM datasets

The purpose of this study is to determine how the spatial scale of measurement influences ecosystem service valuation. Two land cover datasets were compared: one classified from 1-km imagery and one classified from 30-m imagery. The coarse resolution biome dataset used in this study (called the International Geosphere Biosphere Programme (IGBP) Dataset) was classified from 1-km NOAA-AVHRR imagery and includes 17 biome types. The finer resolution National Land Cover Dataset (NLCD) used in this study was classified from 30-m Landsat Thematic Mapper imagery and has 21 land-cover classes. A common land-cover classification scheme containing eight land-cover types was developed in order to compare the two datasets. The areal extent of these land-cover types in each dataset was determined and then multiplied by the value of the ecosystem services to arrive at a total value for ecosystem services. Generally, the areal extent of Lakes/Rivers, barren areas, urban areas, and wetlands in the NLCD showed the largest increases when compared to their extents in the IGBP dataset. The total value of ecosystem services for every state except New Mexico increased using the NLCD. The total value of ecosystem services for the conterminous US increased by almost 200%. The total value according to the I km resolution IGBP data was 259 billion/yr whereas the total value according to the finer resolution (30 m) NLCD data was over $773 billion/yr. Most of the increase in ecosystem service value can be attributed to the increased extent of wetlands in the NLCD. It is also interesting to note that the total value of ecosystem services in the conterminous US is only 8% of gross domestic product of those states ($8.6 trillion). These methods use land cover as a proxy measure of ecosystem service. Some of the pitfalls and promise of this assumption are discussed in the context of spatially explicit remotely sensed image data. (C) 2002 Elsevier Science B.V. All rights reserved.