Pedometric mapping

Last updated

Pedometric mapping, or statistical soil mapping , is data-driven generation of soil property and class maps that is based on use of statistical methods. [1] [2] [3] Its main objectives are to predict values of some soil variable at unobserved locations, and to access the uncertainty of that estimate using statistical inference i.e. statistically optimal approaches. From the application point of view, its main objective is to accurately predict response of a soil-plant ecosystem to various soil management strategies—that is, to generate maps of soil properties and soil classes that can be used for other environmental models and decision-making. It is largely based on applying geostatistics in soil science, and other statistical methods used in pedometrics.

Contents

Although pedometric mapping is mainly data-driven, it can also be largely based on expert knowledge—which, however, must be utilized within a pedometric computational framework to produce more accurate prediction models. For example, data assimilation techniques, such as the space-time Kalman filter, can be used to integrate pedogenetic knowledge and field observations. [4]

In the information theory context, pedometric mapping is used to describe the spatial complexity of soils (information content of soil variables over a geographical area), and to represent this complexity using maps, summary measures, mathematical models and simulations. [5] Simulations are a preferred way of visualizing soil patterns, as they represent their deterministic pattern (due to the landscape), geographic hot-spots, and short range variability (see image, below).[ citation needed ]

Pedometrics

Pedometrics is the application of mathematical and statistical methods to the study of the distribution and genesis of soils. [6]

The term is a portmanteau of the Greek roots pedos (soil) and metron (measurement). Measurement, in this case, is restricted to mathematical and statistical methods as it relates to pedology, the branch of soil science that studies soil in its natural setting.

Pedometrics addresses soil-related problems when there is uncertainty due to deterministic or stochastic variation, vagueness and lack of knowledge of soil properties and processes. It relies on mathematical, statistical and numerical methods, and includes numerical approaches to classification to deal with a supposed deterministic variation. Simulation models incorporate uncertainty by adopting chaos theory, statistical distribution, or fuzzy logic.

Pedometrics addresses pedology from the perspective of emerging scientific fields such as wavelets analysis, fuzzy set theory and data mining in soil data modelling applications. Its advance is also linked to improvements in remote and close-range sensing. [7]

Pedometric vs. traditional soil mapping

In traditional soil survey, spatial distribution of soil properties and soil bodies can be inferred from mental models, leading to manual delineations. Such methods can be considered subjective, and it is hence difficult or impossible to statistically assess the accuracy of such maps without additional field sampling. Traditional soil survey mapping also has limitations in a multithematic GIS, related to the fact that is often not consistently applied by different mappers, and is largely manual and difficult to automate. Most traditional soil maps are based on manual delineations of assumed soil bodies, to which soil attributes are then attached. [8] [9] With pedometric mapping, all outputs are based on rigorous statistical computing, and are hence reproducible.

Traditional soil polygon map (left) vs pedometric map -- four simulations of Zinc content in top-soil generated using geostatistical simulations as shown in this sp package gallery (right). Traditional soil map vs pedometric map.png
Traditional soil polygon map (left) vs pedometric map — four simulations of Zinc content in top-soil generated using geostatistical simulations as shown in this sp package gallery (right).

Pedometric mapping is based largely on extensive and detailed covariate layers, such as Digital Elevation Model (DEM) derivatives, remote sensing imagery, climatic, land cover and geological GIS layers and imagery. Its evolution can be closely connected with the emergence of new technologies and global, publicly available data sources such as the SRTM DEM, MODIS, ASTER and Landsat imagery, gamma radiometrics and LiDAR imagery, and new automated mapping methods.

Comparison between traditional and pedometric (data-driven) mapping techniques
Expert/knowledge-driven soil mappingData/technology-driven (pedometric) soil mapping
Target variablesSoil types (soil series)Analytical soil properties
Spatial data modelDiscrete (soil bodies)Continuous/hybrid (quantities / probabilities)
Major inputsExpert knowledge / soil profile descriptionLaboratory data / proximal soil sensing
Important covariatesSoil delineations (photo-interpretation)Remote sensing images, DEM-derivatives
Spatial prediction modelAveraging per polygonAutomated (geo)statistics
Accuracy assessmentValidation of soil mapping units (kappa)Cross-validation (RMSE)
Data representationPolygon maps + attribute tables (2D)Gridded maps (2D/3D) + prediction error map or simulations
Major technical aspectCartographic scaleGrid cell size
Soil sampling strategiesFree survey (surveyor selects sampling locations)Statistical (design/model-based sampling designs)

Pedometric vs. digital soil mapping

Pedometric analyses rely strictly on geostatistics, whereas digital soil mapping uses more traditional soil-mapping concepts not strictly pedometric in nature. Also referred to as predictive soil mapping, [10] digital soil mapping relies on computer-assisted inference of soil properties to produce digital maps of discrete soil types. Pedometric mapping does not produce maps delineating discrete soil types.

Methods

Pedometric mapping methods differ based on the steps of soil survey data processing:

  1. Sampling
  2. Data screening
  3. Preprocessing of soil covariates
  4. Fitting of geostatistical model
  5. Spatial prediction
  6. Cross-validation / accuracy assessment
  7. Visualization of outputs

One of the main theoretical basis for pedometric mapping is the universal model of soil variation: [4] [11]

...where is the deterministic part of soil variation, is the stochastic, spatially auto-correlated part of variation, and is the remaining residual variation (measurement errors, short-range variability etc.) that is also possibly dependent on , but it is not modeled. This model was first introduced by French mathematician Georges Matheron, and has proven the Best Unbiased Linear Predictor for spatial data. One way of using this model to produce predictions or simulations is by regression-kriging (also known as universal kriging). With soil data, the model's deterministic component is often based on the soil forming factors of climate, organism, relief, parent material (lithology), and time. This conceptual model, known as the CLORPT model, was introduced to soil-landscape modelling by Hans Jenny. [2]

A special group of pedometric mapping techniques focus on downscaling spatial information that can be area-based or continuous. Prediction of soil classes is also another subfield of pedometric mapping, where specific geostatistical methods are used to interpolate the factor-types of variables.

Pedometric mapping is also based largely on novel technologies for measuring soil properties, also referred to as digital soil mapping techniques. They include:

Related Research Articles

Geostatistics is a branch of statistics focusing on spatial or spatiotemporal datasets. Developed originally to predict probability distributions of ore grades for mining operations, it is currently applied in diverse disciplines including petroleum geology, hydrogeology, hydrology, meteorology, oceanography, geochemistry, geometallurgy, geography, forestry, environmental control, landscape ecology, soil science, and agriculture. Geostatistics is applied in varied branches of geography, particularly those involving the spread of diseases (epidemiology), the practice of commerce and military planning (logistics), and the development of efficient spatial networks. Geostatistical algorithms are incorporated in many places, including geographic information systems (GIS).

<span class="mw-page-title-main">Support vector machine</span> Set of methods for supervised statistical learning

In machine learning, support vector machines are supervised learning models with associated learning algorithms that analyze data for classification and regression analysis. Developed at AT&T Bell Laboratories by Vladimir Vapnik with colleagues SVMs are one of the most robust prediction methods, being based on statistical learning frameworks or VC theory proposed by Vapnik and Chervonenkis (1974). Given a set of training examples, each marked as belonging to one of two categories, an SVM training algorithm builds a model that assigns new examples to one category or the other, making it a non-probabilistic binary linear classifier. SVM maps training examples to points in space so as to maximise the width of the gap between the two categories. New examples are then mapped into that same space and predicted to belong to a category based on which side of the gap they fall.

Factor analysis is a statistical method used to describe variability among observed, correlated variables in terms of a potentially lower number of unobserved variables called factors. For example, it is possible that variations in six observed variables mainly reflect the variations in two unobserved (underlying) variables. Factor analysis searches for such joint variations in response to unobserved latent variables. The observed variables are modelled as linear combinations of the potential factors plus "error" terms, hence factor analysis can be thought of as a special case of errors-in-variables models.

<span class="mw-page-title-main">Kriging</span> Method of interpolation

In statistics, originally in geostatistics, kriging or Kriging, also known as Gaussian process regression, is a method of interpolation based on Gaussian process governed by prior covariances. Under suitable assumptions of the prior, kriging gives the best linear unbiased prediction (BLUP) at unsampled locations. Interpolating methods based on other criteria such as smoothness may not yield the BLUP. The method is widely used in the domain of spatial analysis and computer experiments. The technique is also known as Wiener–Kolmogorov prediction, after Norbert Wiener and Andrey Kolmogorov.

In statistics, a linear probability model (LPM) is a special case of a binary regression model. Here the dependent variable for each observation takes values which are either 0 or 1. The probability of observing a 0 or 1 in any one case is treated as depending on one or more explanatory variables. For the "linear probability model", this relationship is a particularly simple one, and allows the model to be fitted by linear regression.

In statistics, ordinary least squares (OLS) is a type of linear least squares method for choosing the unknown parameters in a linear regression model by the principle of least squares: minimizing the sum of the squares of the differences between the observed dependent variable in the input dataset and the output of the (linear) function of the independent variable.

In spatial statistics the theoretical variogram, denoted , is a function describing the degree of spatial dependence of a spatial random field or stochastic process . The semivariogram is half the variogram.

In robust statistics, robust regression seeks to overcome some limitations of traditional regression analysis. A regression analysis models the relationship between one or more independent variables and a dependent variable. Standard types of regression, such as ordinary least squares, have favourable properties if their underlying assumptions are true, but can give misleading results otherwise. Robust regression methods are designed to limit the effect that violations of assumptions by the underlying data-generating process have on regression estimates.

Digital soil mapping (DSM) in soil science, also referred to as predictive soil mapping or pedometric mapping, is the computer-assisted production of digital maps of soil types and soil properties. Soil mapping, in general, involves the creation and population of spatial soil information by the use of field and laboratory observational methods coupled with spatial and non-spatial soil inference systems.

<span class="mw-page-title-main">Soil map</span>

Soil map is a geographical representation showing diversity of soil types and/or soil properties in the area of interest. It is typically the end result of a soil survey inventory, i.e. soil survey. Soil maps are most commonly used for land evaluation, spatial planning, agricultural extension, environmental protection and similar projects. Traditional soil maps typically show only general distribution of soils, accompanied by the soil survey report. Many new soil maps are derived using digital soil mapping techniques. Such maps are typically richer in context and show higher spatial detail than traditional soil maps. Soil maps produced using (geo)statistical techniques also include an estimate of the model uncertainty.

In statistics, generalized least squares (GLS) is a method used to estimate the unknown parameters in a linear regression model when there is a certain degree of correlation between the residuals in the regression model. In such cases, ordinary least squares and weighted least squares may need to be more statistically efficient or else lead to misleading inferences. GLS was first described by Alexander Aitken in 1935.

The Richards equation represents the movement of water in unsaturated soils, and is attributed to Lorenzo A. Richards who published the equation in 1931. It is a quasilinear partial differential equation; its analytical solution is often limited to specific initial and boundary conditions. Proof of the existence and uniqueness of solution was given only in 1983 by Alt and Luckhaus. The equation is based on Darcy-Buckingham law representing flow in porous media under variably saturated conditions, which is stated as

The European Soil Database is the only harmonized soil database in Europe from which many other data information and services are derived. For instance, the European Soil Database v2 Raster Library contains raster (grid) data files with cell sizes of 1km x 1km for a large number of soil related parameters. Each grid is aligned with the INSPIRE reference grid. These rasters are in the public domain and allow expert users to use the data for instance to run soil-, water- and air related models.. The European Soil Database may be downloaded from the European Soil Data Center.

The hat operator is a mathematical notation with various uses in different branches of science and mathematics.

Pedodiversity is the variation of soil properties within an area. Pedodiversity studies were first started by analyzing soil series–area relationships. According to Guo et al. (2003) the term pedodiversity was developed by McBratney (1992) who discussed landscape preservation strategies based on pedodiversity. Recently, examinations of pedodiversity using indices commonly used to characterize bio-diversity have been made. Ibáñez et al. (1995) first introduced ecological diversity indices as measures of pedodiversity. They include Species richness, abundance, and Shannon index. Richness is the number of different soil types, which is the number of soil classes at particular level in a taxonomic system. Abundance is defined as the distribution of the number of soil individuals.

In statistics, polynomial regression is a form of regression analysis in which the relationship between the independent variable x and the dependent variable y is modelled as an nth degree polynomial in x. Polynomial regression fits a nonlinear relationship between the value of x and the corresponding conditional mean of y, denoted E(y |x). Although polynomial regression fits a nonlinear model to the data, as a statistical estimation problem it is linear, in the sense that the regression function E(y | x) is linear in the unknown parameters that are estimated from the data. For this reason, polynomial regression is considered to be a special case of multiple linear regression.

In statistics and in machine learning, a linear predictor function is a linear function of a set of coefficients and explanatory variables, whose value is used to predict the outcome of a dependent variable. This sort of function usually comes in linear regression, where the coefficients are called regression coefficients. However, they also occur in various types of linear classifiers, as well as in various other models, such as principal component analysis and factor analysis. In many of these models, the coefficients are referred to as "weights".

In applied statistics and geostatistics, regression-kriging (RK) is a spatial prediction technique that combines a regression of the dependent variable on auxiliary variables with interpolation (kriging) of the regression residuals. It is mathematically equivalent to the interpolation method variously called universal kriging and kriging with external drift, where auxiliary predictors are used directly to solve the kriging weights.

Kernel methods are a well-established tool to analyze the relationship between input data and the corresponding output of a function. Kernels encapsulate the properties of functions in a computationally efficient way and allow algorithms to easily swap functions of varying complexity.

In statistics, linear regression is a linear approach for modelling the relationship between a scalar response and one or more explanatory variables. The case of one explanatory variable is called simple linear regression; for more than one, the process is called multiple linear regression. This term is distinct from multivariate linear regression, where multiple correlated dependent variables are predicted, rather than a single scalar variable.

References

  1. ↑ Hengl, Tomislav (2003). Pedometric mapping : bridging the gaps between conventional and pedometric approaches. [Wageningen: s.n. ISBN   9789058088963.
  2. 1 2 Grunwald, Sabine, ed. (2006). Environmental soil-landscape modeling geographic information technologies and pedometrics. Boca Raton, FL: CRC/Taylor & Francis. ISBN   9780824723897.
  3. ↑ Kempen, B.; Heuvelink, G. B. M.; Brus, D. J.; Stoorvogel, J. J. (10 March 2010). "Pedometric mapping of soil organic matter using a soil map with quantified uncertainty". European Journal of Soil Science. 61 (3): 333–347. doi:10.1111/j.1365-2389.2010.01232.x. S2CID   94372825.
  4. 1 2 Heuvelink, G.B.M; Webster, R (30 April 2001). "Modelling soil variation: past, present, and future". Geoderma. 100 (3–4): 269–301. Bibcode:2001Geode.100..269H. doi:10.1016/S0016-7061(01)00025-8.
  5. ↑ Hengl, T.; Nikolić, M.; MacMillan, R.A. (31 March 2012). "Mapping efficiency and information content". International Journal of Applied Earth Observation and Geoinformation. 22: 127–138. doi:10.1016/j.jag.2012.02.005.
  6. ↑ Heuvelink, Gerard (December 2003). "The Definition of Pedometrics" (PDF). Pedometron. International Working Group on Pedometrics - Provisional Commission on Pedometrics of the International Union of Soil Sciences (15). Archived from the original (PDF) on 2007-09-30. Retrieved 2006-11-01.
  7. ↑ Peter A., Burrough; Johan Bouma; Scott R. Yates (1994). "The state of the art in pedometrics" (PDF). Geoderma. Elsevier Science B. V., Amsterdam. 62 (1–3): 311–326. Bibcode:1994Geode..62..311B. doi:10.1016/0016-7061(94)90043-4 . Retrieved 2006-11-01.
  8. ↑ McBratney, A.B; Mendonça Santos, M.L; Minasny, B (1 November 2003). "On digital soil mapping". Geoderma. 117 (1–2): 3–52. Bibcode:2003Geode.117....3M. doi:10.1016/S0016-7061(03)00223-4.
  9. ↑ Behrens, Thorsten; Scholten, Thomas (1 June 2006). "Digital soil mapping in Germany—a review". Journal of Plant Nutrition and Soil Science. 169 (3): 434–443. doi:10.1002/jpln.200521962.
  10. ↑ Scull, P.; J. Franklin; O.A. Chadwick; D. McArthur (June 2003). "Predictive soil mapping - a review". Progress in Physical Geography. Sage Publications. 27 (2): 171–197. doi:10.1191/0309133303pp366ra. S2CID   787741.
  11. ↑ Burrough, P.A. (1986). Principles of Geographical Information Systems for Land Resources Assessment. Clarendon Press. p. 194.
This page is based on this Wikipedia article
Text is available under the CC BY-SA 4.0 license; additional terms may apply.
Images, videos and audio are available under their respective licenses.