Premium
A Digital Morphometric Approach for Quantifying Ped Shape
Author(s) -
Mohammed Aoesta K.,
Hirmas Daniel R.,
Giménez Daniel,
Mandel Rolfe D.,
Miller James R.
Publication year - 2016
Publication title -
soil science society of america journal
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.836
H-Index - 168
eISSN - 1435-0661
pISSN - 0361-5995
DOI - 10.2136/sssaj2016.06.0203
Subject(s) - morphometrics , computer science , shape analysis (program analysis) , heuristic , artificial intelligence , categorical variable , pattern recognition (psychology) , machine learning , biology , ecology , static analysis , programming language
Core Ideas Ped shapes were quantified from digital photographs. Circularity was the most useful morphometric for distinguishing ped types. Experience and training influenced ability to recognize prismatic peds. Existing photographs can be analyzed with this approach without the need to resample. Ped shape is an important property with considerable influence over soil processes, such as root penetration, water infiltration, and solute transport. Despite the host of methods employed to quantify other soil morphological properties, ped shape quantification remains elusive. Existing methods attempting to quantify soil structure utilize laboratory techniques that have limitations on sample size and resolution. Our goal was to overcome these limitations by developing an approach to quantify ped shape using morphometrics created from published digital photographs of soil profiles and structure specimens. In addition, ped shapes from heuristic diagrams and three‐dimensional (3‐D) scans of peds were examined. The heuristic diagrams were quantified to assess ped shapes derived from common conceptualizations of soil structure, while the 3‐D scans were quantified to assess the effect of ped orientation on shape measurements. Ped shape was quantified by manually outlining distinct examples of soil peds from high‐resolution photographs and heuristic diagrams and calculating several morphometrics from the resulting silhouettes using image analysis software. A survey was designed to poll expert judgment to properly classify the shape of these peds. Using this method, we were able to transform typical categorical and subjective descriptions of peds into continuous quantitative shape data. The shape metrics, circularity and width to height ratio, exemplify the type of continuous variables that allow significant differences between ped shapes to be detected. This approach opens the door to analyzing soil structure at regional and continental scales through the analysis of existing photographs without the need to resample.