3rd World Congress on GIS and Remote Sensing
Charlotte, USA
Chizumba Shepande
University of Zambia, Zambia
Title: Designing a geospatial analysis tool for inventory and mapping of soils
Biography
Biography: Chizumba Shepande
Abstract
A soil inventory of the Chongwe region of Zambia was prepared using computer aided analysis of Landsat 7 ETM to determine the feasibility of Landsat data in soil mapping and how accurate soil spectral maps produced by digital analysis of satellite imagery can be and how such maps can improve the quality of soil survey in developing countries like Zambia. Thirteen spectral classes were produced. Overall, the work shows that there is a good agreement between Landsat Spectral data and field observation data with a classification accuracy of 72%. This is an indication that there is a definite relationship between Landsat imagery and soil types.
This work shows that visual interpretation and digital analysis of Landsat 7 ETM images have the capacity to map soils with reasonable accuracy. It also demonstrates the capability of Landsat 7 data to delineate soil patterns, especially when acquired during the dry season when there are long periods of cloud-free skies, low soil moisture and minimum vegetation cover.
The work also attempted to determine how the soil spectral data produced by digital analysis of Landsat 7 imagery compared with field observation data. Spectral classes of soil are correlated with individual soil types at sub-group level for all mapping units in the study area of interest. In situations where there is poor agreement between Landsat 7 data and field observation data, possible causes of such discrepancies are determined and explained.