328-10 Estimating Soil Clay Content in Situ with a Visnir Penetrometer Fore Optic.

See more from this Division: SSSA Division: Soil Physics
See more from this Session: Sensors and Instrumentation for Mapping and Monitoring Applications: I
Tuesday, November 4, 2014: 3:40 PM
Long Beach Convention Center, Room 101A
Share |

Matteo Poggio, David Brown and Ross S. Bricklemyer, Washington State University, Pullman, WA
There is growing interest in developing sensors and techniques capable of rapid and inexpensive in situ soil characterization. Several researchers proved the efficiency of visible and near infrared (VisNIR) diffuse spectroscopy in laboratory as technique to investigate soil properties. To further reduce costs and increase spatial resolution of soil data, it would be beneficial employing VisNIR diffuse spectroscopy directly in the field. In this study, we evaluate in situ performances of new VisNIR penetrometer fore optic, capable of simultaneously collecting soil spectra and insertion force. Using hydraulic push-type soil coring systems for insertion (e.g. Giddings), we interrogated soil profiles with the VisNIR penetrometer at 11 fields and 248 sub-field locations throughout the US Pacific Northwest region to a depth of 80 cm, collecting intact cores at each location to build spectral model. 3-cm depth increments were extracted at field interrogation depths for lab clay % determination, with dried and sieved samples also scanned with a commonly used instrument, ASD Contact Probe (Analytical Spectral Devices Inc., Boulder, CO, USA) in the lab. Using partial least squares regression and core-out cross-validation, in situ VisNIR spectra yielded cross-validation results (SEP = 5.2%) comparable with spectra obtained from dried/ground samples from ASD contact probe. Localizing predictions with field-specific calibrations improved predictions (SEP = 4.7%). To improve clay content estimation, we collected soil insertion force profile with the VisNIR penetrometer, at 161 sub-field locations used previously for soil spectra interrogation (5 fields). Preliminary results indicated correlation between soil insertion data and bulk density profile, suggesting the use of both tip resistance data and VisNIR spectra to efficiently characterize soil profile in the field.
See more from this Division: SSSA Division: Soil Physics
See more from this Session: Sensors and Instrumentation for Mapping and Monitoring Applications: I
<< Previous Abstract | Next Abstract