HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text Free Full Text (PDF) Free Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in ISI Web of Science Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via ISI Web of Science (2) Citing Articles via Google Scholar Google Scholar Articles by Nemes, A. Articles by van Genuchten, M. Th. Search for Related Content PubMed Articles by Nemes, A. Articles by van Genuchten, M. Th. GeoRef GeoRef Citation Agricola Articles by Nemes, A. Articles by van Genuchten, M. Th. Related Collections Soil Hydrology Pedotransfer Functions Soil Physics

Sensitivity Analysis of the Nonparametric Nearest Neighbor Technique to Estimate Soil Water Retention

^a Univ. of California, Dep. of Environmental Sciences, Riverside, CA 92521
^b USDA-ARS Hydrology and Remote Sensing Lab., 10300 Baltimore Ave., Bldg. 007, BARC-West, Beltsville, MD 20705
^c USDA-ARS Environmental Microbial Safety Lab., Powder Mill Road, Bldg. 173, BARC-East, Beltsville, MD 20705
^d USDA-ARS George E. Brown Jr. Salinity Lab., 450 West Big Springs Rd., Riverside, CA 92501

View larger version (24K): [in this window] [in a new window]   Fig. 1. Distribution of samples in the NRCS, HYPRES, and BRAZ data sets according to the NRCS textural triangle.

View larger version (22K): [in this window] [in a new window]   Fig. 2. Estimation accuracy using three different weighting methods to make estimations for the NRCS (below) and HYPRES (above) test data sets.

View larger version (27K): [in this window] [in a new window]   Fig. 3. Running root mean squared residuals (RMSR) for the (a) NRCS and (b) HYPRES test data sets for up to 200 ensembles using sand, silt, clay, bulk density, and organic matter content as input and water retention at –1500 kPa matric potential as output.

View larger version (22K): [in this window] [in a new window]   Fig. 4. Correlation between the absolute values of the estimation errors and the actual distance values (di) of each of the k neighbors. Different data set sizes are averaged.

View larger version (18K): [in this window] [in a new window]   Fig. 5. Correlation between the absolute values of the estimation errors and the actual distance values (di) of each of the k neighbors. Different data set sizes are expanded for the NRCS data set for estimating water retention at –1500 kPa matric potential.

View larger version (22K): [in this window] [in a new window]   Fig. 6. Correlation between the absolute values (ABS) of the estimation errors and the distance values (di) of the sixth neighbor. Estimations are shown for estimating water retention at –1500 kPa matric potential for the NRCS test data set using 100 samples in the reference data set and all described inputs (sand, silt, clay content, bulk density, and organic matter content).

View larger version (19K): [in this window] [in a new window]   Fig. 7. Sand and organic matter (OM) contents in the limited NRCS data subset used in Case Study 5. Data points symbolized by open circles (N = 2) are considered hypothetical outliers.

View larger version (18K): [in this window] [in a new window]   Fig. 8. Mean root mean squared residual (RMSR) values and their standard deviations for estimating water retention at (above) –33 kPa and (below) –1500 kPa matric potential obtained using different scenarios to include outliers in the data. Outlier settings: 1 = outliers always in the reference data; 2 = outliers mixed in randomly; 3 = outliers always in the test data; 4 = outliers not in the data. Vertical bars represent ±1 standard deviation, based on 200 ensembles. SSC = sand, silt, and clay content; SSCBD = sand, silt, clay, and bulk density; SSCOM = sand, silt, clay, and organic matter contents; SSCBDOM = sand, silt, clay, and organic matter contents and bulk density.

View larger version (26K): [in this window] [in a new window]   Fig. 9. Running input attribute weights to minimize estimation root mean squared residuals (RMSR) of (a) water retention at –33 kPa and (b) –1500 kPa matric potential. The model used sand, silt, and clay content, bulk density (Db), and organic matter (OM) content as input and 1600 samples in the reference data set.