Research Article

Use of Airborne Hyperspectral Imagery to Map Soil Properties in Tilled Agricultural Fields

Table 2

Partial least squares (PLS) prediction model goodness of fit ( ) associated with each of 15 math treatments, for the 13 analytes that predicted with , calculated using data from all 315 soil sampling locations1. The first derivative gap two (1Dg2, depicted in bold) was selected as the overall most successful model although it was occasionally outperformed (italic). See Section 2 for analyte descriptions.

DerivativeGapCSandSiltClaypHOMKCaMgMnZnFeAlaverage

NON00.5780.7620.7630.5850.4420.6850.5550.6690.6920.6420.6700.7540.7770.659
1ST10.5550.7700.7610.5650.5170.7060.5220.6540.7190.6240.6060.7070.7990.654
1ST20.5910.7630.7630.6170.5490.7170.5140.6760.7080.6380.6470.7400.7820.670
1ST40.5950.7540.7400.5960.4510.6920.5130.6300.6160.5650.6750.7030.7730.639
1ST80.5840.7480.7390.6000.4130.6680.5780.6350.6420.5800.6360.6370.7270.630
1ST160.5880.7400.7440.5500.4420.6720.5470.6100.6190.5670.5910.6660.7250.620
1ST320.5250.7300.7220.5540.3810.6400.5200.6480.6190.5420.5800.6690.7650.607
1ST640.3380.6230.6190.4700.3170.5370.4130.5190.5800.4880.5680.6360.6690.521
2ND10.5420.5850.5970.4240.3610.5260.4840.7080.7050.5290.5990.6650.7290.573
2ND20.5070.6810.6820.3720.4990.5660.4930.6650.6640.5660.5910.6730.7690.595
2ND40.5760.7140.7240.4090.4490.6600.4840.6720.6690.6060.6130.7490.7520.621
2ND80.5360.6980.6890.5330.4720.6020.5280.5950.6310.5420.6380.7430.7530.612
2ND160.5800.7290.7250.5730.4870.6610.4980.5900.6400.5980.6170.6800.7120.622
2ND320.5640.7150.6980.6090.4320.6310.5480.6220.5920.5480.6060.6390.6880.607
2ND640.5250.5960.5890.4210.1560.5990.5110.5560.4640.5010.5380.6020.6140.513

1Analytes that predicted poorly (<0.5) included: (<0.303), (<.355), Cu (<0.358), B (<0.169), S (<0.282), and P saturation (<0.127).