No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
New approaches to the exploitation of former soil survey data
Abstract
A number of soil data is available from the soil and land resource surveys carried out in the past. Modern mathematical and statistical methods and computer development enable new analyses and exploitation of these data. This is required by more complex understanding of soil functions, danger of soil degradation, different agricultural practices, etc. This contribution shows some examples of new exploitation of soil data resulting from a soil survey done in the 1960's. Selected region is characterized by high heterogeneity and variability of soils and natural conditions in general. Characteristics (pH, CEC, organic carbon content, texture, etc.) of more than 600 profiles of agricultural soils were used. Different pedometric methods were applied to evaluate these data. Those included: 1) geostatistics for the analysis of soil spatial dependence and kriging for spatial prediction; 2) multivariate statistical and geostatistical methods for the analysis of interrelations among soil properties and for the determination of principal factors controlling soil heterogeneity in the region; 3) numerical classification for the delineation of soil categories defined according to different objectives (classical soil categorisation, soil vulnerability to pollution by risk elements, suitability to agricultural production, a.o.); 4) pedotransfer functions for modelling the behaviour of chemicals in soils. The results of numerical classification were compared to the results of traditional soil classification. The technology of geographical information systems was used for processing of the resulting applied maps. It was shown that the data of the former soil survey may be used for current purposes using modern methods of their processing. New soil surveys should be therefore focused mainly on special tasks on limited areas, verification of the older data, and validation of the results of mathematical simulation and prediction.
... uses hidden units to extract useful information from inputs and use them to predict outputs Carré and Girard (2002), McBratney et al. (2003), Moonjun et al. (2010), Jafari et al. (2013), Bodaghabadi et al. (2015)Fuzzy systems Indicates uncertainty in predictive and predicted traits or classes. Using fuzzy logic, they draw a certain input to an expected outputZhu et al. (1997),Boruvka et al. (2002),Carré and Girard (2002),Rizzo et al. (2016),Mohamed (2020) ...
Land management is the most challenging issue in the evolving world today, especially in urban areas where most people live. As the population grows, it must be planned to yield enough energy and food, and preserve some species (plants and animals) on Earth. Thus, the obvious need for mapping and efficient soil management in meeting the growing request for food and other basic requirements of communities worldwide. Mapping the spatial distribution of soil formation parameters is very important and serious for information on soil management decisions. Soil formation parameters are different in terms of presence and composition in each landscape, so it is necessary that all information on land resources are collected so that land planners can obtain land use with the maximum yield from each plot of land. Soil maps allow for studying adequate information on soil types and their distribution, and it is required for a variety of planning and usage, including urban planning, forestry, agriculture, grazing, watershed management and engineering. Digital soil mapping (DSM) is a subfield of soil science that requires three components: input data for field and laboratory observations, the process used in terms of spatial and non-spatial soil inference systems, and the output in the form of rasters of forecast along with forecast uncertainty. Digital mapping approaches using data mining and environmental covariates involve climate, organisms, relief, parent material, and time or age of soil which can be determined in soil mapping. The advancement of DSM is due to the confluence of several factors, including the increased availability of spatial data (satellite imagery), computing power for data processing, and the improvement of data mining tools and GIS and geostatistics. This chapter examines the history of soil mapping, environmental covariates, role of remote sensing in DSM, soil inference systems and quality assessments in the DSM. Soil data is needed to address agriculture, natural resource management, development of urban areas and urbanism issues such as water and food security, climate change, land degradation and biodiversity management. Therefore, an updated and accurate soil information are demanded in the studies of international organisations.
... Some soil types may have many pedon sites while others may have a few. These limits the applications of statistical and geostatistical methods which are commonly used in digital soil mapping (Myers, 1994;Boruvka et al., 2002;Diem and Comrie, 2002;Penížek and Borůvka, 2004;Lagacherie, 2008). An alternative is a similarity-based approach (Holt, 1999;Zhu et al., 2010;Pla et al., 2013), which makes soil predictions at unsampled locations based on environmental similarities of the sampled locations. ...
Conventional soil maps generally contain one or more soil types within a single soil polygon. But their geographic locations within the polygon are not specified. This restricts current applications of the maps in site-specific agricultural management and environmental modelling. We examined the utility of legacy pedon data for disaggregating soil polygons and the effectiveness of similarity-based prediction for making use of the under- or over-sampled legacy pedon data for the disaggregation. The method consisted of three steps. First, environmental similarities between the pedon sites and each location were computed based on soil formative environmental factors. Second, according to soil types of the pedon sites, the similarities were aggregated to derive similarity distribution for each soil type. Third, a hardening process was performed on the maps to allocate candidate soil types within the polygons. The study was conducted at the soil subgroup level in a semi-arid area situated in Manitoba, Canada. Based on 186 independent pedon sites, the evaluation of the disaggregated map of soil subgroups showed an overall accuracy of 67% and a Kappa statistic of 0.62. The map represented a better spatial pattern of soil subgroups in both detail and accuracy compared to a dominant soil subgroup map, which was commonly used in practice. Incorrect predictions mainly occurred in the agricultural plain area and the soil subgroups that are very similar in taxonomy, indicating that new environmental covariates need to be developed. We concluded that the combination of legacy pedon data with similarity-based prediction is an effective solution for soil polygon disaggregation.
... (Penizek and Boruvka, 2004). That means that quite often two profiles were located very close to each other, but represented different soils, and therefore they were quite different (Boruvka et al., 2003). This leads to very high nugget in the variograms of soil properties, which certainly caused problems for reasonable interpolation. ...
Present development in precision agriculture has increased the need for reliable and comprehensive information concerning the land surfaces. Traditional soil maps often fail to provide the high-resolution soil data that is now required for sustainable land-management practices and there is a stronger need to understand soil variation at a finer scale than previously demanded. The context of this project is to predict soil attributes with a better resolution than provided by the existing soil maps. Electromagnetic conductivity and gamma-ray emission from the soils can be assessed with airborne sensors allowing generation of maps with substantially more details than existing maps. But there is the need to combine these data and evaluate which remotely sensing data are best suited for different soil properties and the extent to which field data needs to be collected.
In the first part of the study I estimated the scales of variability of soil properties from mapped, surveyed and imaged information sources using geostatistical tools, in the second part I evaluated relationships between soil properties and terrain and geophysics data using regression analysis.
This study has shown that airborne radiometric data clearly depicts true pattern of soil variability. The maps that were produced using conventional soil mapping correspond extremely well with pattern in the gamma emissions. The conventional soil survey data cannot be used in present state to interpolate soil properties at a high degree of spatial accuracy. The existing sampling density did not enable detecting short-scale soil variability.
The chosen terrain attributes cannot be used effectively in the prediction of the selected soil properties at Angas Bremer area due to plain relief and low vertical accuracy of used DEM.
Regression analysis shows the potential in using geophysical data in soil mapping. Of the data analysed, particle size exhibited the strongest correlation with radiometrics.
... However, all interpolation techniques, geostatistical or other, require a fairly dense network of sampling sites from which data are collected. The amount of the data is very often limited due to the cost of collecting soil attributes (Kalivas et al. 2002, Borůvka et al. 2003, Wu et al. 2003. Another limitation, when data from conventional soil survey are used, can be their availability and exploitability due to a specific density and distribution of sampling sites over the space. ...
The objective of this study was to investigate the benefits of methods that incorporate terrain attributes as covariates into the prediction of soil depth. Three primary terrain attributes – elevation, slope and aspect – were tested to improve the depth prediction from conventional soil survey dataset. Different methods were compared: 1) ordinary kriging (OK), 2) co-kriging (COK), 3) regression-kriging (REK), and 4) linear regression (RE). The evaluation of predicted
results was based on comparison with real validation data. With respect to means, OK and COK provided
the best prediction (both 110 cm), RE and REK gave the worst results, their means were significantly lower (79 and 108 cm, respectively) than the mean of real data (111 cm). F-test showed that COK with slope as covariate gave the best result with respect to variances. COK also reproduced best the range of values. The use of auxiliary terrain data improved the prediction of soil depth. However, the improvement was relatively small due to the low correlation of the primary variable with used terrain attributes.
... The sites were chosen to describe different soil units. That means that quite often two profiles were located very close to each other, but represented different soils, and therefore they were quite different (Boruvka et al., 2003). This leads to very high nugget in the variograms of soil properties, which certainly caused problems for reasonable interpolation. ...
Soil is a fundamental natural resource; it is the basis of human agriculture (Scull P. et al., 2003). Present development in sustainable land-management and precision agriculture has increased the need for reliable and comprehensive information concerning the land surface. Traditional soil maps often fail to provide the high-resolution soil data that is now required for sustainable land-management practices (Tunstall, 1998) and there is a stronger need to understand soil variation at a finer scale than previously demanded (Rampant and Abuzar, 2004).
... After soil maps are published, the soil – landscape relationships are sometimes published as journal articles (e.g., Beckmann et al., 1974; Chen, 1997). Sometimes, soil surveys are used to gain insight into soil properties and their distributions (e.g., Breuwsma et al., 1986; Bouma et al., 1996; Boruvka et al., 2002). Soil scientists claim that soil maps can be used for a variety of purposes related to land management but many other potential users have wondered, ''What is the use of soil maps? ...
Soil survey leads to the publication of soil maps and textual reports to communicate the knowledge gained by the surveyors. The argument that soil maps and their legends are representations of structured knowledge is made and discussed with some examples. An epistemology of soil survey is outlined. The relationship between the mental model of a soil surveyor, a soil map legend, and a soil map is formally represented. An analogy between knowledge engineers and soil scientists is made to understand how artificial intelligence, logic, and its formalism can be used in soil survey. Some frameworks for presenting soil surveyors' mental models explicitly are suggested to improve soil survey reporting in the 21st century.
We review various recent approaches to making digital soil maps based on geographic information systems (GIS) data layers, note some commonalities and propose a generic framework for the future. We discuss the various methods that have been, or could be, used for fitting quantitative relationships between soil properties or classes and their ‘environment’. These include generalised linear models, classification and regression trees, neural networks, fuzzy systems and geostatistics. We also review the data layers that have been, or could be, used to describe the ‘environment’. Terrain attributes derived from digital elevation models, and spectral reflectance bands from satellite imagery, have been the most commonly used, but there is a large potential for new data layers. The generic framework, which we call the scorpan-SSPFe (soil spatial prediction function with spatially autocorrelated errors) method, is particularly relevant for those places where soil resource information is limited. It is based on the seven predictive scorpan factors, a generalisation of Jenny’s five factors, namely: (1) s: soil, other or previously measured attributes of the soil at a point; (2) c: climate, climatic properties of the environment at a point; (3) o: organisms, including land cover and natural vegetation; (4) r: topography
Soil grid data were gathered from 110 points and three depths in the Vagia plain, Greece. A total of 48 soil parameters for each point were studied and reduced using multivariate analysis techniques, based on a common variance criterion, and Principal Component Analysis, without loosing less than 8% of the topological information. They were finally reduced to twelve properties which were the most important: electrical conductivity in the third sampling depth (EC3), pH in the third sampling depth (PH3), liquid limit of the third sampling depth (LL3), salt content in the third sampling depth (SALT3), clay content in the third sampling depth (C3), sand content in the third sampling depth (S3) etc. With this reduced set of variables, soils were grouped in seven distinct groups by means of clustering techniques. The numerical clustering showed a congruence of 79.6% compared with the classification based on Soil Taxonomy at the order level. Interpolation by means of kriging was tried both for the first principal component and for individual soil parameters. The spherical model was fitted to the estimated semivariogram, and the range was found to be 2700 m offering a guide for the density of soil sampling and a-posteriori confirmation of the adopted grid size. So using grid data and numerical variable reduction techniques, reliable soil maps can be produced cost-effectively, while this approach is suggested to be applied and evaluated in other areas.