|Lee, Kyou Seung - Sung Kyun Kwan University|
|Sudduth, Kenneth - Ken|
|Lee, Dong Hoon - Sung Kyun Kwan University|
|Chung, Sun Ok - Chungnam National University|
Submitted to: Transactions of the ASABE
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 5/17/2010
Publication Date: 6/29/2010
Citation: Lee, K., Sudduth, K.A., Drummond, S.T., Lee, D., Kitchen, N.R., Chung, S. 2010. Calibration Methods for Soil Property Estimation Using Reflectance Spectroscopy. Transactions of the ASABE. 53(3):675-684.
Interpretive Summary: Measuring the variation in soil properties within fields is an important component of precision agriculture. For many soil properties, it is difficult to obtain enough data to accurately characterize their spatial variation, due to the cost of traditional sampling and laboratory analysis. Sensors that can estimate soil properties without the need for sampling are a promising alternative. One technology that has received considerable attention in this regard is optical reflectance sensing in the visible and near infrared (or NIR) wavelength bands. An important question to answer in applying reflectance sensing for soil analysis is how many, and what type of, calibration soil samples are required. To help answer that question, we collected multiple soil samples from ten fields in five Midwestern states and measured their reflectance characteristics in the laboratory. We used statistical techniques to relate the reflectance to laboratory-measured soil properties. We found that it was important to include calibration soil samples that had characteristics (such as soil type) closely related to the soils under analysis. This could be done either with a very broad, built-in calibration or by creating modified calibrations specific to the soils being studied. Two methods of creating these modified calibrations were studied, and each would have advantages in certain circumstances. These results provide information that instrumentation engineers and researchers can use to help develop and apply new in-field soil sensing technology.
Technical Abstract: Optical diffuse reflectance sensing is a potential approach for rapid and reliable on-site estimation of soil properties. One issue with this sensing approach is whether additional calibration is necessary when the sensor is applied under conditions (e.g., soil types or soil moisture conditions) different from those used to generate an initial calibration, and if so, how many sample points are required in this additional calibration. In this study, these issues were addressed using data from 10 fields from 5 states in the U.S. Corn Belt. Partial least squares (PLS) regression was used to develop calibrations between soil properties and reflectance spectra. Model evaluation was based on the ratio of standard deviation to RMS error (RPD), a statistic commonly used in spectral analysis. When sample data from the field where calibrations were to be applied (i.e., test field) were included in the calibration stage (full information calibration), RPD values of prediction models were increased by an average of 0.55 (from 1.08 to 1.63) compared with results from models not including data from the test field (calibration without field-specific information). Including some samples from the test field (hybrid calibration) generally increased RPD to 90% of that from full information calibration (average increase = 0.49) by using data from 8 to 20 soil cores, with little further improvement given additional data. Using test field points as a bias adjustment (two-stage calibration) increased RPD by an average of 0.29 with 2 to 6 sample points, a finding that was confirmed by Monte Carlo simulation. These results show the importance of including in a calibration set samples similar (i.e., obtained from the same or similar fields) to those in the test set. These similar samples could be included directly in the calibration, or could be used to implement a post-calibration bias adjustment. Although results were more accurate with the recalibration approach, the bias adjustment approach was more efficient computationally and required less data. Thus, either might be preferred depending on specific circumstances.