Lansat-5 Thematic Mapper (Landsat-5 TM) images are evaluated for their use to upgrade a land-cover/land-use database in Switzerland. The Hectometric Information Grid or Hectaraster consists of a database, covering all of Switzerland according to 12 Level-I landcover/land-use categories. In this study Level-I and Level-II “build-up” and “forest” classes are computed using Landsat-5 TM spectral data of the study area of Switzerland. A procedure has been developed that automatically determines the dominant land-cover/land-use inside each hectare or “hectel” of the Hectaraster. The newly created Hectaraster results in a more accurate representation of the actual land-cover/land-use situation in the Nyon test area, as compared to the traditional 1972 Hectaraster. Also, more detail is available, since Level-II categories have been added to the Level-I land-cover/land-use categories.

The study indicates how classified Landsat-5 TM images can efficiently contribute to the management and planning of low density build-up zones. The integration of classified remote-sensing data into a geographic information system (GIS) allows the extraction of new land-use categories by examining their spatial relationships. Thus, in addition to built-up categories, computed from automated classification procedures, potential construction sites and nonused construction areas have been determined through GIS operations.

Methods are presented, which allow the classification of forest stands according to their surface and their shape. The perpendicular vegetation index (PVI) is computed for deciduous and coniferous forest stands. Deciduous and coniferous forests are studied separately. Coniferous forests revealed essentially low PVI values. Dense deciduous forest had the highest PVI values, while open deciduous and deciduous/coniferous stands had intermediate PVI values.