Table 3: Calibration and validation accuracies for approach A—Multiple linear regression of spectral parameters. The models further to be applied for a large-scale prediction of soil constituents based on HyMap imagery are highlighted.

Spectral datasetNo. of spectral variablesCalibration (941/1232 samples)Validation (311/402 samples)
RMSCal RPDCal RMSValRPDVal

ASD spectral resolution(1) In situ field12 of 160.770.472.090.490.541.26
(2) Bare soil field14 of 160.790.442.220.560.481.42
(3) Laboratory13 of 160.750.472.010.740.361.93
(1) In situ field19 of 210.630.801.640.240.901.07
Iron oxides(2) Bare soil field15 of 210.640.781.680.210.931.04
(3) Laboratory19 of 210.750.762.010.231.630.69
(1) In situ field14 of 160.314.121.200.015.130.88
Clay(2) Bare soil field11 of 160.214.391.130.034.800.94
(3) Laboratory16 of 160.234.601.140.054.521.00

HyMap spectral resolution(1) In situ field13 of 160.790.442.190.510.541.27
(2) Bare soil field12 of 160.810.432.280.620.431.57
(3) Laboratory15 of 160.770.462.080.770.352.00
(1) In situ field18 of 210.620.801.640.171.040.93
Iron oxides(2) Bare soil field19 of 210.660.761.730.260.931.03
(3) Laboratory20 of 210.730.791.940.251.580.71
(1) In situ field14 of 160.284.201.180.005.050.89
Clay(2) Bare soil field14 of 160.174.491.110.014.910.92
(3) Laboratory12 of 160.254.531.160.054.620.98

1Number of samples in training and test set for field datasets. 2Number of samples in training and test set for laboratory datasets.