Random Forests Applied as a Soil Spatial Predictive Model in Arid Utah

We sought to predict soil classes by applying random forests (RF), a decision tree analysis, to predict 24 soil classes across an arid watershed of western Utah. Environmental covariates were derived from Landsat 7 Enhanced Thematic Mapper Plus (ETM+) and digital elevation models (DEM). Random forests are similar to classification and regression trees (CART). However, RF is doubly random. Many (e.g., 500) weak trees are grown (trained) independently because each tree is trained with a new randomly selected bootstrap sample, and a random subset of variables is used to split each node. To train and validate the RF trees, 561 soil descriptions were made in the field. An additional 111 points were added by case-based reasoning using photo interpretation. As RF makes classification decisions from the mode of many independently grown trees, model uncertainty can be derived. Furthermore, the probability that a pixel belongs to one or more classes in the legend can be determined. The overall out of the bag (OOB) error for discrete classes was 55.2%. The confusion matrix revealed that four soils that frequently co-occurred on landforms were frequently misclassified as each other. These soils were combined into six soil map units. To identify pixels that might belong to one of these newly created combinations of soil classes, minimum threshold probabilities were set. Employing probability by class can be an effective and objective method of determining membership in soil map unit associations and complexes mapped at the 1:24,000 scale.

[1]  M. Yehia,et al.  In situ reflectance measurements and TM data of some sedimentary rocks with emphasis on white sandstone, southwestern Sinai, Egypt , 1994 .

[2]  D. Soller,et al.  Divisions of geologic time - Major chronostratigraphic and geochronologic units , 2009 .

[3]  B. Hudson,et al.  The Soil Survey as Paradigm-based Science , 1992 .

[4]  Anònim Anònim Keys to Soil Taxonomy , 2010 .

[5]  Elisabeth N. Bui,et al.  Spatial data mining for enhanced soil map modelling , 2002, Int. J. Geogr. Inf. Sci..

[6]  Todd R. Lookingbill,et al.  An empirical approach towards improved spatial estimates of soil moisture for vegetation analysis , 2004, Landscape Ecology.

[7]  A-Xing Zhu,et al.  A case-based reasoning approach to fuzzy soil mapping , 2002 .

[8]  Michael N. DeMers,et al.  Fundamentals of Geographic Information Systems , 1996 .

[9]  R. V. Rossel,et al.  Soil organic carbon prediction by hyperspectral remote sensing and field vis-NIR spectroscopy: An Australian case study , 2008 .

[10]  Lehi F. Hintze,et al.  Geologic History of Utah , 1988 .

[11]  Bill Fiero Geology of the Great Basin , 1986 .

[12]  Lake Bonneville New Data on the Isostatic Deformation of , 1963 .

[13]  G. Pope Soils and Geomorphology, 3rd edn , 2000 .

[14]  John K. Williams,et al.  USING RANDOM FORESTS AND FUZZY LOGIC FOR AUTOMATED STORM TYPE IDENTIFICATION , 2008 .

[15]  M. Baumgartner,et al.  Remote sensing and geographic information systems , 1996 .

[16]  Frederik P. Agterberg,et al.  Interactive spatial data analysis , 1996 .

[17]  P. Jungerius Soils and Geomorphology , 1985 .

[18]  Norbert Silvera,et al.  Accuracy of interpolation techniques for the derivation of digital elevation models in relation to landform types and data density , 2006 .

[19]  Philippe Lagacherie,et al.  Addressing Geographical Data Errors in a Classification Tree for Soil Unit Prediction , 1997, Int. J. Geogr. Inf. Sci..

[20]  D. M. Lemmon,et al.  Geologic map of the Milford Quadrangle and east half of the Frisco Quadrangle, Beaver County, Utah , 1989 .

[21]  Larry Biehl,et al.  Use of combined digital elevation model and satellite radiometric data for regional soil mapping , 2000 .

[22]  John R. Jensen,et al.  Introductory Digital Image Processing: A Remote Sensing Perspective , 1986 .

[23]  B. Minasny,et al.  On digital soil mapping , 2003 .

[24]  Budiman Minasny,et al.  From pedotransfer functions to soil inference systems , 2002 .

[25]  H. Jenny,et al.  The Soil Resource , 1982, Ecological Studies.

[26]  C. Brungard Alternative sampling and analysis methods for digital soil mapping in southwestern Utah , 2009 .

[27]  B. Henderson,et al.  Australia-wide predictions of soil properties using decision trees , 2005 .

[28]  Holly Croft,et al.  Remote sensing of soil surface properties , 2009 .

[29]  Johannes R. Sveinsson,et al.  Random Forests for land cover classification , 2006, Pattern Recognit. Lett..

[30]  J. L. Boettinger,et al.  Chapter 27 Pedogenic Understanding Raster Classification Methodology for Mapping Soils, Powder River Basin, Wyoming, USA , 2006 .

[31]  P. Chavez Image-Based Atmospheric Corrections - Revisited and Improved , 1996 .

[32]  W. Dietrich,et al.  The soil production function and landscape equilibrium , 1997, Nature.

[33]  J. Boettinger,et al.  Chapter 28 Incorporating Classification Trees into a Pedogenic Understanding Raster Classification Methodology, Green River Basin, Wyoming, USA , 2006 .

[34]  R. D. Ramsey,et al.  Digitally Mapping Gypsic and Natric Soil Areas Using Landsat ETM Data , 2007 .

[35]  R. M. Wallace,et al.  Terrain Analysis Using Digital Elevation Models , 2001 .

[36]  J. Noller Applying Geochronology in Predictive Digital Mapping of Soils , 2010 .

[37]  Leo Breiman,et al.  Random Forests , 2001, Machine Learning.

[38]  D. Tarboton A new method for the determination of flow directions and upslope areas in grid digital elevation models , 1997 .

[39]  Jack Major,et al.  Factors of Soil Formation , 1981 .

[40]  Paul E. Gessler,et al.  Soil-Landscape Modelling and Spatial Prediction of Soil Attributes , 1995, Int. J. Geogr. Inf. Sci..

[41]  Peter Scull,et al.  Predictive soil mapping: a review , 2003 .

[42]  A. Zhu Mapping soil landscape as spatial continua: The Neural Network Approach , 2000 .

[43]  J. Peters,et al.  Random forests as a tool for ecohydrological distribution modelling , 2007 .

[44]  P. Scull,et al.  The application of classification tree analysis to soil type prediction in a desert landscape , 2005 .

[45]  Paul E. Gessler,et al.  Modeling Soil–Landscape and Ecosystem Properties Using Terrain Attributes , 2000 .

[46]  Alex B. McBratney,et al.  Soil pattern recognition with fuzzy-c-means : application to classification and soil-landform interrelationships , 1992 .