Vol. 4, No. 1, 9-17, 2005

Abstract
Agricultural productivity is determined and limited in general by a combination of the natural environment and technical measures. If non-capital intensive management is assumed, the natural potential and constraints are of specific importance for the agricultural land use and its productivity. In this context, the Potsdam Institute for Climate Impact Research (PIK) in cooperation with the Max Planck Institute for Meteorology have developed an indicator for natural agricultural potential or rather natural marginal agricultural sites, the so-called marginality index for agricultural land use on a global scale. The aim of its development was to determine the agricultural potential and to calculate the threat of environmental degradation due to agricultural land use. The index analyses several environmental factors limiting agricultural production under low capital input. But global data with a spatial resolution of 0.5°´0.5° can give only a very general idea about spatial distribution and degree of agricultural marginality. Therefore, same influencing factors but with higher spatial resolution and an adapted fuzzy logic based algorithm were used to calculate the agricultural potential within an iterative process with a spatial resolution of 0.05° for Western Africa focusing on the country of Benin. Beyond this spatial aspect, the applicability of remote sensing data within the approach was also studied. The approach comes out with very encouraging results at a regional scale, proving that valuable information can be derived using remote sensing based data with higher spatial resolution.

ISSN 1729-3782