Premium
Scope to map soil management units at the district level from remotely sensed γ ‐ray spectrometry and proximal sensed EM induction data
Author(s) -
Jing Y. Z.,
Huang J.,
Banks R.,
Triantafilis J.
Publication year - 2017
Publication title -
soil use and management
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.709
H-Index - 81
eISSN - 1475-2743
pISSN - 0266-0032
DOI - 10.1111/sum.12359
Subject(s) - topsoil , environmental science , digital soil mapping , soil survey , soil map , remote sensing , subsoil , cluster analysis , scale (ratio) , soil science , computer science , hydrology (agriculture) , soil water , statistics , cartography , mathematics , geology , geography , geotechnical engineering
Abstract The traditional method of soil mapping involves classifying soil into pre‐existing classes using morphological observations and then air‐photograph interpretation to extrapolate the information. To accelerate the process, less costly ancillary data can be used to assist mapping. However, digital soil mapping ( DSM ) is still affected by the classifications used to identify soil types. One reason is because the morphological characteristics are not mutually exclusive, which causes misclassification. In this study, we used a DSM approach, where ancillary data were surrogate for morphological data, with soil types identified by numerical clustering of remotely and proximally sensed data collected across a farming district near Gunnedah, Australia. Remotely sensed data were obtained from an air‐borne gamma‐ray ( γ ‐ray) spectrometer survey, including potassium (K), thorium (Th), uranium (U) and total counts. Proximally sensed data were measured using EM 38 (i.e. EM 38h and EM38v). Using fuzzy k ‐means and a linear mixed model with measured physical (e.g. clay) and chemical (e.g. CEC ) properties from the topsoil (0–0.30 m) and subsoil (0.9–1.2 m), we found that k = 5 was also optimal given that mean‐squared prediction error (i.e. σ p , C 2 ) was minimised. The approach highlighted subtle differences in physical and chemical properties in productive areas. The DSM was unsuccessful in identifying small units; however, inclusion of elevation data might overcome this limitation. This research has implications for providing fast, accurate and meaningful DSM at a district scale, where traditional methods are too expensive.