List of Figures
Figure 1. Study area showing part of Uttar Pradesh state of
India, covering the districts of Varanasi, Chanduali, Sant Ravidas Nagar
and Mirzapur, localized between 82°30′ and 83°30′ East and 24°30′ and
25°30′ North.
Figure 2. Quantile- Quantile plot for (a) electrical
conductivity (untransformed) and (b) log10 transformed
electrical conductivity
Figure 3. (a) Reflectance Spectra in the visual- near infrared
(Vis-NIR) range (350-2500 nm) of soils.
Figure 3. (b) Reflectance Spectra in the middle infra-red (MIR)
range of soils.
Figure 4 (a). Correlation between electrical conductivity and
reflectance in the visual- near infrared (Vis-NIR) region
Figure 4 (b). Correlation between EC and reflectance in the
middle infra-red (MIR) region
Figure 5 . Calibration model developed for EC in the NIR region
using (a) Partial Least Square Regression (PLSR) (b) Random Forest
Regression (RF) (c) Support Vector Regression (SVR) and (d) Multivariate
Adaptive Regression Splines (MARS)
Figure 6 . Validation model developed for EC in the NIR region
using (a) Partial Least Square Regression (PLSR) (b) Random Forest
Regression (RF) (c) Support Vector Regression (SVR) and (d) Multivariate
Adaptive Regression Splines (MARS)
Figure 7 . Calibration model developed for EC in the MIR region
using (a) Partial Least Square Regression (PLSR) (b) Random Forest
Regression (RF) (c) Support Vector Regression (SVR) and (d) Multivariate
Adaptive Regression Splines (MARS)
Figure 8 . Validation model developed for EC in the MIR region
using (a) Partial Least Square Regression (PLSR) (b) Random Forest
Regression (RF) (c) Support Vector Regression (SVR) and (d) Multivariate
Adaptive Regression Splines (MARS)