Topsoil thickness prediction at the catchment scale by integration of invasive sampling, surface geophysics, remote sensing and statistical modeling

Summary Topsoil thickness is a critical input in hydrological modeling because it controls, in conjunction with soil hydraulic properties, the partitioning of water fluxes between the atmosphere and the subsurface. To parameterize a distributed hydrological model that computes groundwater recharge, we developed a data-integration method to predict the clayey topsoil thickness (CTT) that we applied in a small catchment in Portugal (∼19 km 2 ). The prediction method is based on the integration of: (i) invasive sampling used as a CTT reference dataset (61 invasive measurements); (ii) surface geophysics applied to complement the time-consuming invasive sampling; (iii) remote sensing (RS) image processing (high resolution QuickBird image, aerial photographs and ASTER GDEM) used to derive soils classes and terrain parameters; (iv) geostatistical mixed linear model (MLM) applied to integrate the CTT variability at the catchment scale using geophysical and RS derived auxiliary variables. The selection of the appropriate statistical model derived from the MLM was based on the verification of model assumptions using diagnostic tools. We first converted 436 Geonics™ EM-31 field measurements of soil apparent electrical conductivity (EC a ) into CTT. This was achieved by building MLM based calibration models that integrated 25 invasive CTT measurements paired with corresponding EC a and RS-derived auxiliary variables. Next, we predicted the CTT at the catchment scale by applying the MLM approach and integrating the RS-derived auxiliary variables with: (i) the 436 CTT values derived from surface geophysical dataset; (ii) the 61 CTT values from the reference invasive dataset. The two maps had similar CTT patterns which depicted the spatial variability of the CTT over the geomorphologic catchment features. The prediction map derived from the geophysical dataset resulted in slightly lower CTT values than the reference map (median of 0.87 m against 1.11 m) and a comparable accuracy (RMSE of 0.76 m against 0.88 m). As these differences will be minimized during the calibration process of the hydrological model, the presented methodology is considered suitable for hydrological and environmental studies, in which catchments often need to be investigated over large areas.

[1]  A. McBratney,et al.  Further results on prediction of soil properties from terrain attributes: heterotopic cokriging and regression-kriging , 1995 .

[2]  Ioannis Z. Gitas,et al.  The development of an object-oriented classification model for operational burned area mapping on the Mediterranean island of Thasos using LANDSAT TM images. , 2002 .

[3]  Okke Batelaan,et al.  GIS-based recharge estimation by coupling surface–subsurface water balances , 2007 .

[4]  D. P. Shrestha,et al.  Digital terrain analysis in ILWIS : lecture note , 2003 .

[5]  Edzer Pebesma,et al.  Mapping groundwater quality in the Netherlands , 1997 .

[6]  Mark S. Seyfried,et al.  Dielectric Loss and Calibration of the Hydra Probe Soil Water Sensor , 2005 .

[7]  John Triantafilis,et al.  COMPARISON OF STATISTICAL PREDICTION METHODS FOR ESTIMATING FIELD-SCALE CLAY CONTENT USING DIFFERENT COMBINATIONS OF ANCILLARY VARIABLES , 2001 .

[8]  Mark S. Seyfried,et al.  Temperature Effects on Soil Dielectric Properties Measured at 50 MHz , 2007 .

[9]  G. Quinn,et al.  Experimental Design and Data Analysis for Biologists , 2002 .

[10]  Dennis L. Corwin,et al.  Prediction of spatial soil property information from ancillary sensor data using ordinary linear regression: Model derivations, residual assumptions and model validation tests , 2008 .

[11]  D. J. Brus,et al.  The use of electromagnetic measurements of apparent soil electrical conductivity to predict the boulder clay depth , 1992 .

[12]  John Triantafilis,et al.  Five Geostatistical Models to Predict Soil Salinity from Electromagnetic Induction Data Across Irrigated Cotton , 2001 .

[13]  Michael Sommer,et al.  Mapping Clay Content across Boundaries at the Landscape Scale with Electromagnetic Induction , 2007 .

[14]  Sanford Weisberg,et al.  An R Companion to Applied Regression , 2010 .

[15]  B. Maathuis,et al.  Digital Elevation Model Based Hydro‐processing , 2006 .

[16]  K. Rushton,et al.  Improved soil moisture balance methodology for recharge estimation , 2006 .

[17]  Mark S. Seyfried,et al.  Measurement of soil water content with a 50-MHz soil dielectric sensor , 2004 .

[18]  D. Corwin,et al.  Apparent soil electrical conductivity measurements in agriculture , 2005 .

[19]  R. Reese.  Geostatistics for Environmental Scientists , 2001 .

[20]  J. Finch.  Estimating direct groundwater recharge using a simple water balance model – sensitivity to land surface parameters , 1998 .

[21]  J. Triantafilis,et al.  Mapping clay content variation using electromagnetic induction techniques , 2005 .

[22]  D. Brus,et al.  A comparison of kriging, co-kriging and kriging combined with regression for spatial interpolation of horizon depth with censored observations , 1995 .

[23]  Spatio-temporal Groundwater Recharge Assessment:A Data-integration and Modelling Approach , 2008 .

[24]  U. Benz,et al.  Multi-resolution, object-oriented fuzzy analysis of remote sensing data for GIS-ready information , 2004 .

[25]  Edzer J. Pebesma,et al.  Multivariable geostatistics in S: the gstat package , 2004, Comput. Geosci..

[26]  Tomislav Hengl,et al.  A Practical Guide to Geostatistical Mapping , 2009 .

[27]  Julian J. Faraway,et al.  Practical Regression and Anova using R , 2002 .

[28]  A. Stein,et al.  Soil sampling strategies for spatial prediction by correlation with auxiliary maps , 2003 .

[29]  M. Lubczynski,et al.  Integration of various data sources for transient groundwater modeling with spatio-temporally variable fluxes—Sardon study case, Spain , 2005 .

[30]  Kenneth A. Sudduth,et al.  Comparison of electromagnetic induction and direct sensing of soil electrical conductivity , 2003 .

[31]  Dennis L. Corwin,et al.  Characterizing soil spatial variability with apparent soil electrical conductivity , 2005 .

[32]  David G. Rossiter,et al.  Prediction of soil depth using environmental variables in an anthropogenic landscape, a case study in the Western Ghats of Kerala, India. , 2009 .

[33]  Biswajeet Pradhan,et al.  Improvement of land-use classification using object-oriented and fuzzy logic approach , 2009 .

[34]  Clemens Reimann,et al.  Statistical data analysis explained : applied environmental statics with R , 2008 .

[35]  Maciek W. Lubczynski,et al.  Spatio-temporal groundwater recharge assessment using a lumped - parameter distributed model of the unsaturated zone, pyEARTH-2D. , 2010 .

[36]  Dominique King,et al.  Using multiple external drifts to estimate a soil variable , 2003 .

[37]  J. Fox.  Nonparametric Regression Appendix to An R and S-PLUS Companion to Applied Regression , 2002 .

[38]  F. Husson,et al.  Statistique avec R , 2008 .

[39]  D. Corwin,et al.  Application of Soil Electrical Conductivity to Precision Agriculture , 2003 .

[40]  Alex B. McBratney,et al.  A comparison of prediction methods for the creation of field-extent soil property maps , 2001 .

[41]  Shmulik P. Friedman,et al.  Soil properties influencing apparent electrical conductivity: a review , 2005 .

[42]  Kenneth A. Sudduth,et al.  Estimating depths to claypans using electromagnetic induction methods , 1994 .

[43]  Andrew D. Weiss.  Topographic position and landforms analysis , 2001 .

[44]  Dominique King,et al.  Comparison of kriging with external drift and simple linear regression for predicting soil horizon thickness with different sample densities. , 2000 .

[45]  Korbinian Strimmer,et al.  APE: Analyses of Phylogenetics and Evolution in R language , 2004, Bioinform..