Land Cover Classification System for the Lower Delta of the Parana River (Argentina): Its Relationship with Landsat Thematic Mapper Spectral Classes

P. Kandus, H. Karszenbaum, L. Frulla

Abstract


The Lower Delta of The Parana River is a tidal freshwater wetland. The land cover characteristics of the Lower Delta of the Parana River were determined by its hydrologic regime, the geomorphological patterns of the islands, and the human activity developed in the region. This paper addresses the development of a conceptual land cover classification scheme and its relationship with classes derived from Landsat/TM satellite imagery. First, a specific conceptual land cover classification scheme is presented based on the definition of delimitation and characterization variables. The regional conceptual classification scheme developed was intended to facilitate comparisons with other systems, and to simplify the use of satellite imagery for land cover map production. Next, the resulting "information classes" were associated to "spectral" classes using a layered classification approach based on the application of an unsupervised classification procedure to a set of multitemporal Landsat/TM imagery. In each step, the spectral classes were subject to careful labeling and recoding based on field work and aerial photography. The final land cover map obtained is the result of combining the different thematic layers obtained. The resulting overall classification accuracy was on the order of 83%. A thorough discussion of the results is presented, where some of the characteristics of this complex area are highlighted together with the advantages and disadvantages of using satellite imagery for this type of product.

Keywords


Satellite remote sensing; vegetation mapping; coastal ecosystems.

Full Text:

PDF

Refbacks

  • There are currently no refbacks.