Shoot-out (2004)

Designing the Best Possible Calibration Procedure for
for Mid-Infrared Diffuse Reflectance Spectra

Jim Reeves, Coordinator of ShootOut-2004 Session

Objectives of the Shootout: Spectra of 569 samples are provided from the far near-infrared to the mid-infrared spectral region. All spectra were obtained at 4 wavenumber resolution using diffuse reflectance on an FTIR and will be the basis of the ShootOut at the next Chambersburg meeting (August 8-13, 2004). All spectra are in pseudo-absorbance (KBr background used) and were obtained originally as SPC files and converted to the formats presented by either myself or Howard Mark. Other formats may be available per participant request. The values for four analytes for each sample are also provided. It is up to the individual to determine how to split or not split the dataset into calibration, validation or test files, as part of examining differing views of calibration development, and to determine if all or only part of the spectral range is needed or advisable. The overall objective being to develop the best method for calibration development for these near-/mid-infrared diffuse reflectance spectra..

James B. Reeves, III, USDA-ARS, Beltsville, MD