TY - JOUR
T1 - Visible near infrared diffuse reflectance spectroscopy (VisNIR DRS) for rapid measurement of organic matter in compost
AU - McWhirt, Amanda L.
AU - Weindorf, David C.
AU - Chakraborty, Somsubhra
AU - Li, Bin
PY - 2012/10
Y1 - 2012/10
N2 - Commercial compost is the inherently variable organic product of a controlled decomposition process. In the USA, assessment of compost's physicochemical parameters presently relies on standard laboratory analyses set forth in Test Methods for the Examination of Composting and Compost (TMECC). A rapid, field-portable means of assessing the organic matter (OM) content of compost products would be useful to help producers ensure optimal uniformity in their compost products. Visible near infrared diffuse reflectance spectroscopy (VisNIR DRS) is a rapid, proximal-sensing technology proven effective at quantifying organic matter levels in soils. As such, VisNIR DRS was evaluated to assess its applicability to compost. Thirty-six compost samples representing a wide variety of source materials and moisture content were collected and scanned with VisNIR DRS under moist and oven-dry conditions. Partial least squares (PLS) regression and principal component regression (PCR) were used to relate the VisNIR DRS spectra with laboratory-measured OM to build compost OM prediction models. Raw reflectance, and first- and second-derivatives of the reflectance spectra were considered. In general, PLS regression outperformed PCR and the oven-dried first-derivative PLS model produced an r2 value of 0.82 along with a residual prediction deviation value of 1.72. As such, VisNIR DRS shows promise as a suitable technique for the analysis of compost OM content for dried samples.
AB - Commercial compost is the inherently variable organic product of a controlled decomposition process. In the USA, assessment of compost's physicochemical parameters presently relies on standard laboratory analyses set forth in Test Methods for the Examination of Composting and Compost (TMECC). A rapid, field-portable means of assessing the organic matter (OM) content of compost products would be useful to help producers ensure optimal uniformity in their compost products. Visible near infrared diffuse reflectance spectroscopy (VisNIR DRS) is a rapid, proximal-sensing technology proven effective at quantifying organic matter levels in soils. As such, VisNIR DRS was evaluated to assess its applicability to compost. Thirty-six compost samples representing a wide variety of source materials and moisture content were collected and scanned with VisNIR DRS under moist and oven-dry conditions. Partial least squares (PLS) regression and principal component regression (PCR) were used to relate the VisNIR DRS spectra with laboratory-measured OM to build compost OM prediction models. Raw reflectance, and first- and second-derivatives of the reflectance spectra were considered. In general, PLS regression outperformed PCR and the oven-dried first-derivative PLS model produced an r2 value of 0.82 along with a residual prediction deviation value of 1.72. As such, VisNIR DRS shows promise as a suitable technique for the analysis of compost OM content for dried samples.
KW - Visible near-infrared diffuse reflectance spectroscopy (VisNIR DRS)
KW - compost
KW - loss-on-ignition (LOI)
KW - organic matter
UR - http://www.scopus.com/inward/record.url?scp=84867080874&partnerID=8YFLogxK
U2 - 10.1177/0734242X12450601
DO - 10.1177/0734242X12450601
M3 - Article
C2 - 22677915
AN - SCOPUS:84867080874
SN - 0734-242X
VL - 30
SP - 1049
EP - 1058
JO - Waste Management and Research
JF - Waste Management and Research
IS - 10
ER -