Soil mapping approach in GIS using Landsat satellite imagery and DEM data

African journal of agricultural research (Impact Factor: 0.26). 12/2009; 4(11):1295-1302.


The objective of this study was to create base soil survey maps of the studied lands using Landsat satellite imagery and Digital Elevation Model (DEM) data in a GIS framework. Specific goals were to generate soil maps and to test the usage probability of slope class map overlies colour composite images as a preliminary map for soil survey in a hilly terrain. Surrogate soil-landscape data layers were derived from Landsat satellite imagery and a 10 m DEM. The data were also used to produce 3D-view with slope class boundaries superimposed Landsat image and relief shaded map as a colour map in order to select possible site of soil profile pits and to define physiographic units. Six soil series formed on two different physiographic units were determined, described and sampled. Soil profiles have been classified according to Soil Taxonomy and FAO-Unesco soil map of the World legend classification systems. The methodology was adequate for soil survey and mapping of some types of soils.

  • Source
    • "Ongoing research in digital soil mapping indicates that quantitative prediction models are promising tools to produce soil maps with acceptable accuracy [14] [15]. Research has provided optimistic results, and some researchers obtained better results than traditional soil surveys [16] [17] [18]. The use of satellite data to complement topographic information improves the mapping of natural resources [10] [13] [19]. "
    [Show abstract] [Hide abstract]
    ABSTRACT: Information about the spatial distribution of soil properties is necessary for natural resources modeling; however, the cost of soil surveys limits the development of high-resolution soil maps. The objective of this study was to provide an approach for predicting soil attributes. Topographic attributes and the normalized difference vegetation index (NDVI) were used to provide information about the spatial distribution of soil properties using clustering and statistical techniques for the 56 km 2 Gumara-Maksegnit watershed in Ethiopia. Multiple linear regression models implemented within classified subwatersheds explained 6–85% of the variations in soil depth, texture, organic matter, bulk density, pH, total nitrogen, available phosphorous, and stone content. The prediction model was favorably comparable with the interpolation using the inverse distance weighted algorithm. The use of satellite images improved the prediction. The soil depth prediction accuracy dropped gradually from 98% when 180 field observations were used to 65% using only 25 field observations. Soil attributes were predicted with acceptable accuracy even with a low density of observations (1-2 observations/2 km 2 ). This is because the model utilizes topographic and satellite data to support the statistical prediction of soil properties between two observations. Hence, the use of DEM and remote sensing with minimum field data provides an alternative source of spatially continuous soil attributes.
    Full-text · Article · Oct 2013 · Applied and Environmental Soil Science
  • [Show abstract] [Hide abstract]
    ABSTRACT: Lack of topographic details in cadastral maps and limited terrain information in coarse-resolution digital elevation models (DEMs) are limitations in understanding the soil–environment relationship, identification of soil patterns, and their boundaries. The present study describes the technique of cadastral-level soil mapping using Cartosat-1-derived products. A high-resolution DEM with a posting of 10 m generated from a Cartosat-1 stereo pair was used to derive terrain attributes. Based on erosional and depositional processes, five major landforms, namely plateau top, escarpment, pediment (erosional), alluvial plain, and narrow valley (depositional), have been delineated using 3D perspective viewing of the landscape and 10 m contours generated from the DEM. The pediments and alluvial plain were further divided into upper and lower pediments and alluvial plain based on elevation. A detailed slope map has been generated from the DEM and reclassified into seven slope classes, namely nearly level (0–1%), very gently sloping (1–3%), gently sloping (3–5%), moderately sloping (5–10%), strongly sloping (10–15%), moderately steep to steep (15–25%), and steep to very steep (25–50%). Five land-use/land-cover classes, namely single crop, double crop, wasteland with and without scrub, and degraded forest have been delineated using Cartosat-1-sharpened IRS-P6 LISS-IV data. The landforms, slope, and land-use/land-cover maps were integrated in a GIS framework and 45 precise physiography–land-use (PLU) units were derived. Nine soil series were identified in major landforms. Soils of plateau top, escarpment, and pediment associated with a low weathering front are very shallow to shallow in depth, have moderate to severe erosion, slight stoniness, and clay texture, whereas soils of alluvial plain, developed in Deccan basaltic alluvium, are deep to very deep (>150 cm), fine textured with shrink–swell potential, and calcareous in nature. Different phases of soil series were identified based on tone, texture, and pattern variations in the satellite image. A detailed soil map has been developed from the PLU map using soil series information and augur observations through the PLU–soil relationship. The combined use of the high-resolution DEM and satellite data has immensely helped in understanding the soil-forming environment, identification of soil patterns, and their boundaries for precise and faster mapping compared to conventional soil mapping.
    No preview · Article · Apr 2014 · International Journal of Remote Sensing

  • No preview · Chapter · Jul 2015