Digital Terrain Attributes Research Articles

Numerical classification and digital mapping of coastal plain sands of Southeastern Nigeria

The study classified the coastal plain sands of south-eastern Nigeria at the series level and modeled the classification using digital terrain attributes. The study utilized 72 secondary and 12 primary profile pits data generated from 24 and 4 locations (at 3 per location) for classification/modelling and validation respectively. The three profile pits per location represents the three topographic positions of upper, middle and lower slopes. Digital elevation model was also utilized for the generation of terrain attributes. Soil morphological characteristics were coded for suitability in statistical analysis. Hierarchical clustering was utilized in the grouping of the soil into 17 homogeneous groups referred to as soil series. Regression kriging was used to model the predicted soil series within the area covered by coastal plain sands in Akwa Ibom State. The variables that could be used in the modelling of the different classified soil series include Sand Content, aspect, flow accumulation, compound topographic index (CTI), elevation, hill shade, slope, curvature, flow direction, stream power index (SPI), profile curvature, tangential curvature (R2 = 0.21).Out of the 17 soil series classified, 14 was successfully mapped using digital technique. It was observed that 66.7% of the classified soil series were accurately predicted using digital mapping technique. The classifications carried out numerically made use of morphological discrete variables whereas digital used empirically determined continuous variables which could be more accurate. Therefore it could be inferred that the digitally produced soil classification is more accurate and 14 soil series could be identified and mapped in the study area.
 Key words: pedogenesis, digital soil mapping, soil series, hierarchical clusters, regression kriging

Terrain characterization of small island using publicly available data and open- source software: a case study of Marinduque, Philippines

Digital terrain attributes derived from digital elevation model (DEM) such as elevation, slope, and aspect are widely used to determine the influence of topography on different environmental and human processes. The advent of publicly available DEM data has provided a cheaper, low-cost alternative to traditional field data collection and survey. Handling, processing, and visualization of such data on an open-source software will provide researchers and specialists a better and faster way of generating digital terrain maps and creating input data for other analyses. This paper demonstrates the methodology of combining the use ASTER GDEM and SAGA functionality of QGIS, and R software to develop terrain maps for Marinduque, an island province of the Philippines.

Validation of digital soil maps at different spatial supports

In this study we validated digital soil maps of clay content at different spatial supports; point, 48m blocks and soil–land use complexes (SLU). The aim being to examine the change in prediction quality with different prediction supports. Digital soil maps of clay content at depths of 0–10cm, 10–30cm and 30–50cm were created using linear mixed models, legacy soil data and readily available covariates such as digital terrain attributes, landsat and gamma radiometrics. A random stratified sampling design was used to collect an independent validation dataset. There was little change in the accuracy between the point (RMSE=15.7% for 0–10cm, 17.4% for 10–30cm, 15.2% for 30–50cm) and 48m block supports (RMSE=15.5% for 0–10cm, 17.1% for 10–30cm, 13.5% for 30–50cm). However the prediction quality was much improved at the support of the SLUs (RMSE=9.5% for 0–10cm, 10.1% for 10–30cm and 5.1% for 30–50cm). It is a standard practice in digital soil mapping studies to validate at the point support but our results show that this is likely to represent the worst case scenario for assessing their prediction quality. Further work should consider finding optimal prediction supports for creating digital soil maps.

Soil characteristics and herbicide sorption coefficients in 140 soil profiles of two irregular undulating to hummocky terrains of western Canada

Sensitive input parameters such as herbicide sorption coefficients are seldom quantified in sufficient spatial detail, leading to large uncertainties in pesticide fate model predictions at regional and national scales. This study examined the spatial variability of soil characteristics and herbicide sorption coefficients (Kd) in a total of 609 soil samples obtained from 140 soil profiles in two undulating to hummocky soil-landscapes of western Canada. In all soil horizons, Kd values were always the largest for glyphosate followed by atrazine then 2,4-D. In each soil-landscape, SOC, 2,4-D and atrazine Kd values (all strongly associated) numerically decreased with A>B>C mineral horizons, regardless of the segment of the slope in which the soil profile was located. SOC, 2,4-D Kd and atrazine Kd values were typically smaller in the eluvial A-horizons (Ae, Aej) than other A-horizons. In several landform elements, glyphosate Kd values were numerically greater in the B rather than the A or C mineral horizons. Glyphosate Kd values were particularly large in B-horizons with redoximorphic features such as in illuvial B-horizons enriched with clay (Btg, Btgj). When the soil properties and sorption values of similar horizons were compared across the soil-landscapes, SOC and 2,4-D Kd demonstrated greater coefficients of variation than atrazine Kd and glyphosate Kd. The strength of prediction equations for 2,4-D Kd and atrazine Kd were strong in each soil-landscape when all samples in the soil profile were included in the model. However, the strength of prediction equations by mineral horizon was generally weaker and in some cases not significant. The in-field variability of glyphosate Kd values could not be predicted from the measured soil properties data, or calculated digital terrain attributes. Terrain attributes also showed little to no correlation, or predictive strength for describing the variability of 2,4-D and atrazine sorption in the soil-landscapes.

The utility of remotely-sensed vegetative and terrain covariates at different spatial resolutions in modelling soil and watertable depth (for digital soil mapping)

Digital soil modelling and mapping is reliant on the availability and utility of easily derived and accessible covariates. In this paper, the value of covariates derived from a time-series of remotely-sensed ASTER satellite imagery and digital elevation models were evaluated for modelling two soil attributes — soil depth and watertable depth. Modelling was performed at two resolutions: a fine resolution (15m pixels) that relates to the resolution of the ASTER Visible-NIR bands, and a larger resolution (90m pixels) that relates to the resolution of the thermal bands of the ASTER imagery. Upscaling to a larger pixel size and downscaling to a smaller pixel size were performed to adjust the covariates where necessary. A regression tree approach was used to model soil depth and watertable depth, recorded as a binary ‘deep’ or ‘shallow’ response, using the ASTER imagery-derived covariates and digital terrain attributes (DTAs). Modelling was performed at a single spatial resolution (15 or 90m pixels) using the imagery-derived covariates only, the DTAs only, or a mixture of both. A multi-resolution model was also generated, by using both imagery-derived covariates and DTAs at both resolutions. When mixed with the DTAs, the imagery-derived covariates helped explain the uncertainty (variance) in the soil depth data but not in the watertable depth. The ASTER-derived downscaled evapotranspiration-based covariates were of particular significance in the soil depth modelling. Watertable depth was best explained by models that used DTAs at a smaller pixel size. Information on vegetative growth was neither superior nor complementary to information on terrain for modelling watertable depth. Using a multi-resolution model significantly improved the modelling of soil depth but not of watertable depth. The effect of covariate and modelling resolution on model performance is discussed within the context of the GlobalSoilMap.net project.

Characterisation of soil texture variability using the apparent soil electrical conductivity at a highly variable site

The characterisation of the spatial distribution of clay, silt, and sand/gravel is one of the main objectives of soil surveys. Researchers as well as producers have an interest in characterising soil texture variability. The objective of our study was the development of models to derive the clay, silt, and sand/gravel content from the variables apparent electrical conductivity (EC a), the boundary depth between Quaternary and Tertiary sediments, the terrain parameters, and the cultivation (organic vs integrated and type of fertiliser). The investigation site made it possible to include a wide range of soil types within the geological area under investigation. The apparent electrical conductivity and the soil properties texture and digital terrain attributes were densely mapped onto approximately 17 ha. Soil sampling was carried out in a 50⁎50-m grid. Clay and sand/gravel were most closely related to the EC a, whereas silt showed a stronger dependency on the boundary depth. R 2 values ranged between 0.67 and 0.76 in this hilly landscape. However, some weaknesses of the applied procedure were observed: on layered soils with clay lenses covered with sandy, gravelly material, too much clay, and too little sand/gravel were predicted. In some subareas with distinct differences at the field boundaries, breaks in the EC a were observed. The latter was likely due to fertilising effects that caused enhanced EC a levels. In conclusion, the EC a, in combination with the boundary depth between Quaternary and Tertiary sediments, the elevation, the terrain aspect, and the cultivation parameters represent a useful and robust surveying technique to predict soil texture for the Tertiary hill country in southern Germany.

Locating Eroded Waterways with United States Geological Survey Elevation Data

Concentrated water flow causes considerable erosion in agricultural fields. The objective of this study was to test whether a stochastic modeling approach could be used to determine the location of erosion channels based on terrain attributes derived from 10‐m level‐2 U.S. Geological Survey (USGS) elevation data. Logistic regression was used to develop models that predicted where erosion was likely to occur based on digital terrain attributes derived from the elevation data. The model was fit using georeferenced field observations of soil erosion. A simple two‐variable model was developed for predicting erosion channels that included two topographic attributes: the length‐slope factor (t score = 9.7) and the topographic wetness index (t score = 11.4). Leave‐one‐field‐out cross‐validation (average of 22% false positives and 34% false negatives) confirmed that models correctly identified many of the eroded waterway features but did not accurately predict their shapes. This work suggests that erosion indices developed from freely available 10‐m elevation data could be used to help conservation planners identify erosion channels in some agriculture fields. The USGS National Elevation Dataset (NED) provides free, seamless elevation data on 10‐ and 30‐m grid for most of the United States.

Mapping soil drainage classes of Amik Plain using Landsat images

Soil drainage is one of the important soil properties affecting plant growth, water transfer and solute transport in soils. Soil drainage is also an environmental component affecting irrigation and soil reclamation, land capability for agriculture, flood control systems, engineering, health and infectious diseases. The objective of this study was to map soil drainage classes by using Landsat image in Amik Plain (Hatay, Turkey). Terrain and vegetation are characterized by digital terrain attributes, and vegetation indices using a LANDSAT-7 Enhanced Thematic Mapper Image. The study benefits from five data sources: Landsat ETM image, topographic maps, soil maps, State Hydraulic Works (DSI) land cover records and ground data from field surveys. Image classification was carried out using Maximum Likelihood (ML) Classification with supervised training. Soil drainage classes were determined, thus finalizing the process of mapping after each mapping unit and drainage class prepared as a result of the ML classification were validated on site. According to the drainage map prepared using satellite image and ground data, 51,4% (37,234 ha) of Amik Plain are well drained and moderately well drained. 48,6% (35,192 ha) of Amik Plain are somewhat poorly drained, poorly drained, and very poorly drained. Key words: Soil drainage, remote sensing, Landsat images.

An expert system to predict intricate saline - sodic subsoil patterns in upland South Australia

Digital soil mapping (DSM) offers apparent benefits over more labour-intensive and costly traditional soil survey. Large cartographic scale (e.g. 1 : 10 000 scale) soil maps are rare in Australia, especially in agricultural areas where they are needed to support detailed land evaluation and targeted land management decisions. We describe a DSM expert system using environmental correlation that applies a priori knowledge from a key area (128 ha) soil–landscape with a regionally repeating toposequence to predict the distribution of saline–sodic subsoil patterns in the surrounding upland farming region (2275 ha) in South Australia. Our predictive framework comprises interrelated and iterative steps, including: (i) consolidating a priori knowledge of the key area soil–landscape; (ii) refining existing mentally held and graphic soil–landscape models; (iii) selecting suitable environmental covariates compatible with geographic information systems (GIS) by interrogation via 3D visualisation using a GIS; (iv) transforming the existing soil–landscape models to a computer model; (v) applying the computer model to the environmental variables using the expert system; (vi) performing the predictive mapping; and (vii) validation. The environmental covariates selected include: digital terrain attributes of slope gradient, topographic wetness index and plan curvature, and airborne gamma-radiometric K%. We apply selected soil profile physiochemical data from a prior soil survey to validate mapping. Results showed that we correctly predicted the saline–sodic subsoils in 10 of 11 reference profiles in the region.

Analyzing Digital Terrain Attributes to Predict Soil Attributes for a Relatively Large Area

Accurate information about soil attributes, presented in a spatially continuous form, is prerequisite for many land resources management applications. The availability of detailed soil maps and its ability to supply such information for modern tools and applications are questionable. Some alternatives, based on using terrain attributes derived from digital elevation model (DEM) to predict soil attributes are investigated in this study. A study area of 148 km2 in the northern part of Jordan was used. The area is covered by detailed soil map and 2193 field observations, from which the soil attributes were extracted. Terrain attributes derived from 20‐m resolution DEM were utilized to predict soil attributes by implementing different statistical and clustering techniques. The use of multiple linear regression models within small watershed subdivisions enabled the prediction of soil depth for 89.3% of the field observations within ±50 cm, the water‐holding capacity (WHC) for 75.8% within ±50 mm cm−1 and the surface cover percentage for 78.7% within ±10%. The models also predicted surface cover type for 94.5% of the field observations, erosion type for 48.4%, erosion class for 98.0%, and soil texture for 90.3%, within one class difference between predicted and field estimated classes. Comparing these results with estimates of soil attributes using the soil map indicated that the modeling of the soil‐landscape relationships within small watershed subdivisions is a promising approach to predict soil attributes for large areas. An important feature is the spatial distribution of the predicted soil attributes, which is provided in more detailed form than what the soil map provides.

Modeling Soil–Landscape and Ecosystem Properties Using Terrain Attributes

Soil–landscape patterns result from the integration of short‐ and long‐term pedogeomorphic processes. A 2‐ha hillslope catena in California shows short‐distance variation in A horizon depth from 8 to 80 cm and in soil depth from 8 to >450 cm in convex to concave positions. Similar variations in net primary productivity (NPP) and soil C represent significant information often not captured by soil survey maps. Strong correlations between these measured soil–landscape variables and explanatory digital terrain attributes are used to develop quantitative soil–landscape models. We were able to account for between 52 and 88% of soil property variance using easily computed terrain variables such as slope and flow accumulation. Spatial implementation of the models suggest lateral redistribution processes resulting in differential accumulation of C and soil mass in convergent and divergent landscape positions. The models are explicit and quantitative, which enables their use for testing hypotheses about the spatial distribution of fine‐scale landscape and ecosystem processes and for parameterizing spatially distributed hydrological and ecosystem simulation models.