The land-use classification used in KABA is based two different projects and two different sensors. The central part of the study site was classified during an ERS-1 feasibility study called ERSCLIP (ERS-1 CLImate Project; Beha et al., 1996 or Scherer et al., 1996). This multi-sensorial and multi- temporal land-use classification was carried out using one multispectral Landsat-TM image and a data set of three ERS-1 SAR scenes from 1991 and 1992 applying a maximum-likelihood approach. This resulted in 29 Pixel classes geocoded in the Swiss National Coordinate System with a grid resolution of 30m.

The ERSCLIP study site does not cover the whole KABA area. For this reason, a second land-use classification carried out in a project called REKLIP (REgio KLimaProjekt)  was merged with the one of ERSCLIP. The classification done in REKLIP is based on one Landsat-TM image (Scherer et al., 1994) applying also the maximum-likelihood approach. Corresponding subclasses of both land-use classifications were converted into single pixel classes sufficient for the purpose of KABA.

The merging of both land-use classifications resulted in the basic data layer for further work. Areal percentages of these pixel classes of grid elements of 100m and of 25m in a 100m environment were used for the classification of areal types and the calculation of land-use attributed parameters and their distribution in the study site.

For further, more detailed information you may click on Proceedings, see publication list or contact Dr. Dieter Scherer.