Research Article

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

Table 4

Predicted concentrations of select analytes (C, Silt, Al, Fe) derived from pixels of near infrared hyperspectral imagery for each of the six bare soil fields using either unsmoothed (1-pixel) imagery with 2.5 m resolution, or spatially smoothed (9-pixel) imagery to which a low pass filter had been applied1. See Section 2 for analyte descriptions.

CSiltAlFe
Field1-pixel%9-pixel%1-pixel%9-pixel%1-pixel mg/kg9-pixel mg/kg1-pixel mg/kg9-pixel mg/kg

Temple 1Smin0.971.1314.614.4342454056
max1.861.9480.158.71017947474394
mean1.411.4842.141.1633631194192
sd0.100.105.95.098924644

Temple 1Nmin0.850.961.05.6−311473764
max1.952.0468.959.11304897452429
mean1.231.2543.642.1589555200210
sd0.110.115.84.91151126065

Temple 2min−0.480.8114.68.734−3894054
max2.072.9280.175.710171013474529
mean1.271.2842.138.2633512194226
sd0.250.185.97.1981464663

Temple 3min0.390.600.21.2−1370−288−6
max2.292.4880.874.911731031465474
mean1.271.2752.351.7632639173170
sd0.170.167.97.51541457163

Masonmin0.410.44−7.6−7.5−19829000
max2.271.9464.948.512311180391391
mean0.810.8127.728.7755776111108
sd0.110.106.47.1961003227

Schradermin0.910.9812.313.42552965866
max2.082.0561.656.9996971346315
mean1.451.4838.638.3669681177172
sd0.130.156.05.887813534

Overallmin0.971.1314.614.42552965866
max1.861.9461.648.5996897346315
mean1.201.2241.340.5660648168171

1See Table 3 for the accuracy ( ) associated with each analyte prediction model.