Table 5. Top 10 predictors selected by the Cubist calibration model A, B, C and their attribute usage ranking.
| Model A | Attribute usage | Model B | Attribute usage | Model C | Attribute usage |
|---|---|---|---|---|---|
| PAW (%) | 100 % | PAW | 93% | PAW | 100% |
| 1,930 nm | 96% | Landsat8 NDVI(July) | 82% | 1,930nm | 98% |
| Landsat8 NDVI(July) | 88% | Landsat8 NDVI(June) | 77% | Landsat8 NDVI(July) | 90% |
| Landsat8 NDVI(June) | 85% | Landsat8 EVI(July) | 70% | Landsat8 NDVI(June) | 87% |
| Landsat8 EVI(July) | 80% | Landsat8 SR(July) | 63% | Landsat8 LST(July) | 84% |
| Landsat8 EVI(June) | 68% | Landsat8 NDVIgreen (July) | 52% | Landsat8 LST(July) | 75% |
| Landsat8 SR(July) | 56% | Landsat8 EVI(June) | 48% | Landsat8 NDVIgreen (July) | 68% |
| Landsat8 NDVIgreen_July | 40% | Landsat8 SR(June) | 41% | Landsat8 NDVIgreen (June) | 61% |
| SPOT5 SR | 38% | SPOT5 NDVI | 33% | SPOT5 SR | 44% |
| SPOT5 NDVI | 29% | SPOT 5 SR | 26% | Elevation | 22% |
PAW: plant available water; NDVI: normalized differential vegetation index; EVI: Enhanced vegetation Index; SR: Simple Ratio
Model A, upland and wetland model based on ancillary environmental data, remote sensing data and the estimated spectra (1930 nm)
Model B, upland and wetland model based on ancillary environmental data and remote sensing data
Model C, upland model based on ancillary environmental data, remote sensing data and the estimated spectra (1930 nm).