Search

The paper evaluates spreading of observations in feature and geographical spaces as a key to sampling optimisation for spatial prediction by correlation with auxiliary maps. Although auxiliary data are commonly used for mapping soil variables, problems associated with the design of sampling strategies are rarely examined. When generalised least-squares estimation is used, the overall prediction error depends upon spreading of points in both feature and geographical space. Allocation of points uniformly over the feature space range proportionally to the distribution of predictor (equal range stratification, or ER design) is suggested as a prudent sampling strategy when the regression model between the soil and auxiliary variables is unknown. An existing 100-observation sample from a 50 by 50 km soil survey in central Croatia was used to illustrate these concepts. It was re-sampled to 25-point datasets using different experimental designs: ER and 2 response surface designs. The designs were compared for their performance in predicting soil organic matter from elevation (univariate example) using the overall prediction error as an evaluation criterion. The ER design gave overall prediction error similar to the minmax design, suggesting that it is a good compromise between accurate model estimation and minimisation of spatial autucorrelation of residuals. In addition, the ER design was extended to the multivariate case. Four predictors (elevation, temperature, wetness index, and NDVI) were transformed to standardised principal components. The sampling points were then assigned to the components in proportion to the variance explained by a principal component analysis and following the ER design. Since stratification of the feature space results in a large number of possible points in each cluster, the spreading in geographical space can also be maximised by selecting the best of several realisations. Additional keywords: response surface design, soil survey, general least squares, feature space, principal components., Introduction A sampling design in soil survey specifies which points, transects, or areas will be visited for field measurements or observations. Sampling incorporates concepts of survey intensity, spatial variability, and [...]

A method to enhance manual landform delineation using photo-interpretation to map a larger area is described. Conventional aerial photo-interpretation (API) maps using a geo-pedological legend of 21 classes were prepared for six sample areas totaling 111 [km.sup.2] in the Baranja region, eastern Croatia. Nine terrain parameters extracted from a digital elevation model (DEM) (ground water depth, slope, plan curvature, profile curvature, viewshed, accumulation flow, wetness index, sediment transport index, and the distance to nearest watercourse) were used to extrapolate photo-interpretation over the entire survey area (1062 [km.sup.2]). The classification accuracy was assessed using the error matrix, calculated by comparing both the whole API maps and point samples, with the results of classification. The first results, using a maximum-likelihood classifier, were 58.2% (hill land), 39.1% (plain), and 45.3% (entire area) reproducibility of the training set. Six classes in the plain were responsible for a large proportion of the misclassifications, due to an insufficiently detailed DEM and the complex nature of landforms (point bar complexes, levees, active channel banks), which cannot be explained with the terrain parameters only. Reproducibility for a simplified legend of 15 classes over the study area was improved to 65.8% (plain), 58.2% (hill land), and 63.4% (entire area) using the whole-API training set. After the simplification of legend (15) and with the iterative (3) selection of pointsample training set, classification was able to reproduce 97.6% (hill land), 86.7% (plain), and 90.2% (entire area) of the training set. The supervised classification showed fine details not achieved by photo-interpretation. The number of manual photo-interpretations that had to be prepared was reduced from 84 to 6. The methodology can be applied by soil survey teams to edit and update current maps and to enhance or replace API for new surveys.

The accuracy and precision of predictions of soil texture based on visible- near-infrared (Vis-NIR) spectra are affected by the soil moisture content at the time of measurement. This study aimed to quantify these effects. We also developed a method to improve the accuracy of soil texture prediction when the difference in moisture content is large and unknown, as is usually the case in field measurements. Reflection spectra (380-2400 nm) of 89 soil samples were obtained in the laboratory under nine moisture conditions. Prediction models for each moisture condition were built separately using partial least-squares (PLS) regression. Each model was applied to independent validation sets under the same moisture condition as the calibration sets, and then the model based on air-dried soils (dry model) was applied to the other eight moisture conditions to quantify the effect of soil moisture on the prediction accuracy. Finally, the perpendicular drought index (PDI) was used as an indicator of soil moisture, and the nine moisture conditions were regrouped to four groups using the PDI. Prediction models for each group with the same PDI were built and evaluated. The results show that Vis-NIR spectra can be directly applied to predict the soil texture of soils in different known moisture states. When the moisture state is unknown, the models based on PDI grouping markedly improve prediction accuracy. The RMSEp of prediction from PDI-grouped models of prediction of the content of clay and sand separates were between 1 and 2% and between 8 and11%, respectively, for the different moisture classes. This method has potential for direct application in the field with on-the-go sensors. [ABSTRACT FROM AUTHOR]

A method to visualize multiple membership maps, called 'Colour mixture' (CM) is described and compared with alternative techniques: defuzzification and Pixel mixture. Six landform parameters were used to derive the landform classes using supervised fuzzy k -means classification. The continuous categorical map is derived by GIS calculations with colours, where colour values are considered to represent the taxonomic space spanned by the attribute variables. Coordinates of the nine class centres (landform facets) were first transformed from multivariate to two-dimensional attribute space using factor analysis, and then projected on the Hue Saturation Intensity (HSI) colourwheel. The taxonomic value was coded with the Hue and confusion with Saturation. To improve visual impression, saturation was replaced with whiteness. Classes that were closer in attribute space were merged into similar generic colours. The CM technique limits the derived mixed-colour map to seven generic hues independently of the total number of classes, which provides a basis for automated generalization. The confusion index derived from the mixed-colour map was used to derive primary boundaries and to locate areas of higher taxonomic confusion. [ABSTRACT FROM AUTHOR]