Semi-automated disaggregation of a conventional soil map using knowledge driven data mining and random forests in the Sonoran desert, USA

Travis W. Nauman, James A. Thompson, Craig Rasmussen

Research output: Contribution to journalArticle

13 Citations (Scopus)

Abstract

Conventional soil maps (CSM) have provided baseline soil information for land use planning for over 100 years. Although CSM have been widely used, they are not suitable to meet growing demands for high resolution soil information at field scales. We present a repeatable method to disaggregate CSM data into ~30-meter resolution rasterized soil class maps that include continuous representation of probabilistic map uncertainty. Methods include training set creation for original CSM component soil classes from soil-landscape descriptions within the original survey database. Training sets are used to build a random forest predictive model utilizing environmental covariate maps derived from ASTER satellite imagery and the USGS National Elevation Dataset. Results showed agreement at 70 percent of independent field validation sites and equivalent accuracy between original CSM map units and the disaggregated map. Uncertainty of predictions was mapped by relating prediction frequencies of the random forest model and success rates at validation sites.

Original languageEnglish (US)
Pages (from-to)353-366
Number of pages14
JournalPhotogrammetric Engineering and Remote Sensing
Volume80
Issue number4
DOIs
StatePublished - 2014

Fingerprint

data mining
Data mining
desert
Soils
soil
ASTER
prediction
land use planning
satellite imagery
soil map
Satellite imagery
Land use
Planning
method

ASJC Scopus subject areas

  • Computers in Earth Sciences

Cite this

@article{2f925f6cea8f4125b2f4390564c79fa1,
title = "Semi-automated disaggregation of a conventional soil map using knowledge driven data mining and random forests in the Sonoran desert, USA",
abstract = "Conventional soil maps (CSM) have provided baseline soil information for land use planning for over 100 years. Although CSM have been widely used, they are not suitable to meet growing demands for high resolution soil information at field scales. We present a repeatable method to disaggregate CSM data into ~30-meter resolution rasterized soil class maps that include continuous representation of probabilistic map uncertainty. Methods include training set creation for original CSM component soil classes from soil-landscape descriptions within the original survey database. Training sets are used to build a random forest predictive model utilizing environmental covariate maps derived from ASTER satellite imagery and the USGS National Elevation Dataset. Results showed agreement at 70 percent of independent field validation sites and equivalent accuracy between original CSM map units and the disaggregated map. Uncertainty of predictions was mapped by relating prediction frequencies of the random forest model and success rates at validation sites.",
author = "Nauman, {Travis W.} and Thompson, {James A.} and Craig Rasmussen",
year = "2014",
doi = "10.14358/PERS.80.4.353",
language = "English (US)",
volume = "80",
pages = "353--366",
journal = "Photogrammetric Engineering and Remote Sensing",
issn = "0099-1112",
publisher = "American Society for Photogrammetry and Remote Sensing",
number = "4",

}

TY - JOUR

T1 - Semi-automated disaggregation of a conventional soil map using knowledge driven data mining and random forests in the Sonoran desert, USA

AU - Nauman, Travis W.

AU - Thompson, James A.

AU - Rasmussen, Craig

PY - 2014

Y1 - 2014

N2 - Conventional soil maps (CSM) have provided baseline soil information for land use planning for over 100 years. Although CSM have been widely used, they are not suitable to meet growing demands for high resolution soil information at field scales. We present a repeatable method to disaggregate CSM data into ~30-meter resolution rasterized soil class maps that include continuous representation of probabilistic map uncertainty. Methods include training set creation for original CSM component soil classes from soil-landscape descriptions within the original survey database. Training sets are used to build a random forest predictive model utilizing environmental covariate maps derived from ASTER satellite imagery and the USGS National Elevation Dataset. Results showed agreement at 70 percent of independent field validation sites and equivalent accuracy between original CSM map units and the disaggregated map. Uncertainty of predictions was mapped by relating prediction frequencies of the random forest model and success rates at validation sites.

AB - Conventional soil maps (CSM) have provided baseline soil information for land use planning for over 100 years. Although CSM have been widely used, they are not suitable to meet growing demands for high resolution soil information at field scales. We present a repeatable method to disaggregate CSM data into ~30-meter resolution rasterized soil class maps that include continuous representation of probabilistic map uncertainty. Methods include training set creation for original CSM component soil classes from soil-landscape descriptions within the original survey database. Training sets are used to build a random forest predictive model utilizing environmental covariate maps derived from ASTER satellite imagery and the USGS National Elevation Dataset. Results showed agreement at 70 percent of independent field validation sites and equivalent accuracy between original CSM map units and the disaggregated map. Uncertainty of predictions was mapped by relating prediction frequencies of the random forest model and success rates at validation sites.

UR - http://www.scopus.com/inward/record.url?scp=84898469854&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84898469854&partnerID=8YFLogxK

U2 - 10.14358/PERS.80.4.353

DO - 10.14358/PERS.80.4.353

M3 - Article

AN - SCOPUS:84898469854

VL - 80

SP - 353

EP - 366

JO - Photogrammetric Engineering and Remote Sensing

JF - Photogrammetric Engineering and Remote Sensing

SN - 0099-1112

IS - 4

ER -