Algorithms for quantitative pedology: A toolkit for soil scientists

We developed AQP  (Algorithms for Quantitative Pedology) as a software package in the R statistical environment so that we can easily perform common tasks such as visualization, aggregation, and classification of soil profile data.  As soil data can be associated with location (x, y), depth (z), and property space (p);  the high dimensionality and grouped nature of this type of data can complicate standard analysis, summarization, and visualization tasks. The AQP package provides pedometricians with an analysis framework that can handle the complexity of soil profile data.

AQP provides pedometricians with tools for rendering soil profiles graphically, based on horizon boundaries, horizon designation, soil color and soil properties (measured or inferred). Soil profiles can be plotted in a given order, e.g. by the type of landform from which they have been sampled.

Soil property data organised according to genetic horizons are difficult to process due variable horizonation depths. A solution to this problem is offered in AQP by normalizing a collection of horizons, irrespective of the horizon type, according to a common system of ‘‘slices’’. Essentially, each soil property (from each soil profile) is aligned to a common depth basis. With this new data structure it is possible to plot, aggregate, map, or compute numerical measures of similarity by slice.  The use of aggregate depth functions could support a fundamental shift in how soil survey is presented: from the concept of a ‘‘modal profile’’ (i.e. a single pedon) to a collection of ‘‘representative depth functions’’. Representative soil property depth functions would give users a continuous estimate of soil properties and fulfill a long-standing criticism of soil survey regarding the current lack of uncertainty estimates for soil property data.

In addition, the normalization of soil profile horizons allows to run similarity analysis on any given collection of soil profiles. Such similarity measures can then be used in numerical soil classification by leveraging the important base of classification methods published within the extensible and open source R framework. Alternative classification schemes could also be generated from the same underlying data, but directed towards specific goals, by selecting which variables and dissimilarity metrics are used.

Functions in the aqp package have been successfully applied to studies involving several thousand soil profiles. AQP is an open source project. Its scriptable nature allows pedometrics research to be reproducible, but also embeddable in other tools, such as the SoilWeb mobile app.

The stable version of the aqp package is hosted on CRAN, and the development version is hosted on R-Forge. A recent presentation on the “aqp family” of R packages can be found here.

The article is now available as Open-Access until end of the year:

Beaudette, D.E., Roudier, P., O’Geen, A.T., 2013. Algorithms for quantitative pedology: A toolkit for soil scientists. Computers & Geosciences 52, 258–268. doi:10.1016/j.cageo.2012.10.020


You may also like...