The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences
Publications Copernicus
Download
Citation
Articles | Volume XLII-3/W6
https://doi.org/10.5194/isprs-archives-XLII-3-W6-417-2019
https://doi.org/10.5194/isprs-archives-XLII-3-W6-417-2019
26 Jul 2019
 | 26 Jul 2019

VIS-NIR REFLECTANCE SPECTROSCOPY FOR ASSESSMENT OF SOIL ORGANIC CARBON IN A RICE-WHEAT FIELD OF LUDHIANA DISTRICT OF PUNJAB

B. P. Mondal, B. S. Sekhon, R. N. Sahoo, and P. Paul

Keywords: PLSR, Reflectance Spectroscopy, Rice-Wheat, RMSEP, RPD, SOC

Abstract. Soil organic carbon (SOC) is a crucial indicator of soil fertility, maintaining soil health and sustaining the productivity of agro-ecosystem. Rapid, reliable and cost effective assessment of soil properties specially for SOC is important for monitoring soil fertility status along with soil health. Conventional chemical analysis of any soil property is hazardous, tedious and time consuming. So, the visible near infrared (VIS-NIR) reflectance spectroscopy can provide an effective alternative technique for rapid and ecofriendly measurement of soil properties. In view of this, a key soil fertility parameter SOC was examined through diffuse reflectance spectroscopy. Georeferenced surface soil samples (0–15 cm) were collected from a rice-wheat field of the study area for both chemical and spectral analysis. A viable statistical technique namely partial least square regression (PLSR) technique were used to correlate the measured properties with soil reflectance spectra and for developing spectral model. The predictive performance of newly developed spectral model was evaluated through different reliable indices like root mean square of error of prediction (RMSEP), coefficient of determination (R2) and ratio of performance deviation (RPD). The result showed that the R2 value for SOC is 0.44, RMSEP is 0.07 and the RPD value is 1.57 in the validation dataset. The RPD value indicating that SOC can be reliably predicted using the hyperspectral model or reflectance analysis. So, this hyperspectral modeling technique can be successfully employed for monitoring soil health as well as for sustainable agriculture.