Baseline estimates of soil organic carbon by proximal sensing: Comparing design-based, model-assisted and model-based inference

R.A. Viscarra Rossel, D.J. Brus, C. Lobsey, Z. Shi, G. McLachlan

Research output: Contribution to journalArticleAcademicpeer-review

20 Citations (Scopus)

Abstract

For baselining and to assess changes in soil organic carbon (C) we need efficient soil sampling designs and methods for measuring C stocks. Conventional analytical methods are time-consuming, expensive and impractical, particularly for measuring at depth. Here we demonstrate the use of proximal soil sensors for estimating the total soil organic C stocks and their accuracies in the 0-10 cm, 0-30 cm and 0-100 cm layers, and for mapping the stocks in each of the three depth layers across 2837 ha of grazing land. Sampling locations were selected by probability sampling, which allowed design-based, model-assisted and model-based estimation of the total organic C stock in the study area. We show that spectroscopic and gamma attenuation sensors can produce accurate measures of soil organic C and bulk density at the sampling locations, in this case every 5 cm to a depth of 1 m. Interpolated data from a mobile multisensor platform were used as covariates in Cubist to map soil organic C. The Cubist map was subsequently used as a covariate in the model-assisted and model-based estimation of the total organic C stock. The design-based, model-assisted and model-based estimates of the total organic C stocks in the study area were similar. However, the variances of the model-assisted and model-based estimates were smaller compared to those of the design-based method. The model-based method produced the smallest variances for all three depth layers. Maps helped to assess variability in the C stock of the study area. The contribution of the spectroscopic model prediction error to our uncertainty about the total soil organic C stocks was relatively small. We found that in soil under unimproved pastures, remnant vegetation and forests there is good rationale for measuring soil organic C beyond the commonly recommended depth of 0-30 cm.

LanguageEnglish
Pages152-163
JournalGeoderma
Volume265
DOIs
Publication statusPublished - 1 Mar 2016

Fingerprint

soil organic carbon
organic carbon
organic soils
organic soil
soil
sampling
sensors (equipment)
sensor
grazing lands
bulk density
analytical methods
analytical method
pasture
uncertainty
soil sampling
grazing
methodology
pastures
vegetation
prediction

Keywords

  • Design-based sampling
  • Infrared spectroscopy
  • Model-based inference
  • Proximal soil sensing
  • Regression estimator
  • Soil organic carbon stocks
  • Visible-near

Cite this

Viscarra Rossel, R.A. ; Brus, D.J. ; Lobsey, C. ; Shi, Z. ; McLachlan, G. / Baseline estimates of soil organic carbon by proximal sensing : Comparing design-based, model-assisted and model-based inference. In: Geoderma. 2016 ; Vol. 265. pp. 152-163.
@article{b31e5a52ab3f4098be5464460b5669bc,
title = "Baseline estimates of soil organic carbon by proximal sensing: Comparing design-based, model-assisted and model-based inference",
abstract = "For baselining and to assess changes in soil organic carbon (C) we need efficient soil sampling designs and methods for measuring C stocks. Conventional analytical methods are time-consuming, expensive and impractical, particularly for measuring at depth. Here we demonstrate the use of proximal soil sensors for estimating the total soil organic C stocks and their accuracies in the 0-10 cm, 0-30 cm and 0-100 cm layers, and for mapping the stocks in each of the three depth layers across 2837 ha of grazing land. Sampling locations were selected by probability sampling, which allowed design-based, model-assisted and model-based estimation of the total organic C stock in the study area. We show that spectroscopic and gamma attenuation sensors can produce accurate measures of soil organic C and bulk density at the sampling locations, in this case every 5 cm to a depth of 1 m. Interpolated data from a mobile multisensor platform were used as covariates in Cubist to map soil organic C. The Cubist map was subsequently used as a covariate in the model-assisted and model-based estimation of the total organic C stock. The design-based, model-assisted and model-based estimates of the total organic C stocks in the study area were similar. However, the variances of the model-assisted and model-based estimates were smaller compared to those of the design-based method. The model-based method produced the smallest variances for all three depth layers. Maps helped to assess variability in the C stock of the study area. The contribution of the spectroscopic model prediction error to our uncertainty about the total soil organic C stocks was relatively small. We found that in soil under unimproved pastures, remnant vegetation and forests there is good rationale for measuring soil organic C beyond the commonly recommended depth of 0-30 cm.",
keywords = "Design-based sampling, Infrared spectroscopy, Model-based inference, Proximal soil sensing, Regression estimator, Soil organic carbon stocks, Visible-near",
author = "{Viscarra Rossel}, R.A. and D.J. Brus and C. Lobsey and Z. Shi and G. McLachlan",
year = "2016",
month = "3",
day = "1",
doi = "10.1016/j.geoderma.2015.11.016",
language = "English",
volume = "265",
pages = "152--163",
journal = "Geoderma",
issn = "0016-7061",
publisher = "Elsevier",

}

Baseline estimates of soil organic carbon by proximal sensing : Comparing design-based, model-assisted and model-based inference. / Viscarra Rossel, R.A.; Brus, D.J.; Lobsey, C.; Shi, Z.; McLachlan, G.

In: Geoderma, Vol. 265, 01.03.2016, p. 152-163.

Research output: Contribution to journalArticleAcademicpeer-review

TY - JOUR

T1 - Baseline estimates of soil organic carbon by proximal sensing

T2 - Geoderma

AU - Viscarra Rossel, R.A.

AU - Brus, D.J.

AU - Lobsey, C.

AU - Shi, Z.

AU - McLachlan, G.

PY - 2016/3/1

Y1 - 2016/3/1

N2 - For baselining and to assess changes in soil organic carbon (C) we need efficient soil sampling designs and methods for measuring C stocks. Conventional analytical methods are time-consuming, expensive and impractical, particularly for measuring at depth. Here we demonstrate the use of proximal soil sensors for estimating the total soil organic C stocks and their accuracies in the 0-10 cm, 0-30 cm and 0-100 cm layers, and for mapping the stocks in each of the three depth layers across 2837 ha of grazing land. Sampling locations were selected by probability sampling, which allowed design-based, model-assisted and model-based estimation of the total organic C stock in the study area. We show that spectroscopic and gamma attenuation sensors can produce accurate measures of soil organic C and bulk density at the sampling locations, in this case every 5 cm to a depth of 1 m. Interpolated data from a mobile multisensor platform were used as covariates in Cubist to map soil organic C. The Cubist map was subsequently used as a covariate in the model-assisted and model-based estimation of the total organic C stock. The design-based, model-assisted and model-based estimates of the total organic C stocks in the study area were similar. However, the variances of the model-assisted and model-based estimates were smaller compared to those of the design-based method. The model-based method produced the smallest variances for all three depth layers. Maps helped to assess variability in the C stock of the study area. The contribution of the spectroscopic model prediction error to our uncertainty about the total soil organic C stocks was relatively small. We found that in soil under unimproved pastures, remnant vegetation and forests there is good rationale for measuring soil organic C beyond the commonly recommended depth of 0-30 cm.

AB - For baselining and to assess changes in soil organic carbon (C) we need efficient soil sampling designs and methods for measuring C stocks. Conventional analytical methods are time-consuming, expensive and impractical, particularly for measuring at depth. Here we demonstrate the use of proximal soil sensors for estimating the total soil organic C stocks and their accuracies in the 0-10 cm, 0-30 cm and 0-100 cm layers, and for mapping the stocks in each of the three depth layers across 2837 ha of grazing land. Sampling locations were selected by probability sampling, which allowed design-based, model-assisted and model-based estimation of the total organic C stock in the study area. We show that spectroscopic and gamma attenuation sensors can produce accurate measures of soil organic C and bulk density at the sampling locations, in this case every 5 cm to a depth of 1 m. Interpolated data from a mobile multisensor platform were used as covariates in Cubist to map soil organic C. The Cubist map was subsequently used as a covariate in the model-assisted and model-based estimation of the total organic C stock. The design-based, model-assisted and model-based estimates of the total organic C stocks in the study area were similar. However, the variances of the model-assisted and model-based estimates were smaller compared to those of the design-based method. The model-based method produced the smallest variances for all three depth layers. Maps helped to assess variability in the C stock of the study area. The contribution of the spectroscopic model prediction error to our uncertainty about the total soil organic C stocks was relatively small. We found that in soil under unimproved pastures, remnant vegetation and forests there is good rationale for measuring soil organic C beyond the commonly recommended depth of 0-30 cm.

KW - Design-based sampling

KW - Infrared spectroscopy

KW - Model-based inference

KW - Proximal soil sensing

KW - Regression estimator

KW - Soil organic carbon stocks

KW - Visible-near

U2 - 10.1016/j.geoderma.2015.11.016

DO - 10.1016/j.geoderma.2015.11.016

M3 - Article

VL - 265

SP - 152

EP - 163

JO - Geoderma

JF - Geoderma

SN - 0016-7061

ER -