Table 4.
Coefficient of determination and RMSE (from repeated 10-fold cross-validation) for linear regressions between crop traits and vegetation indices calculated from UAV and ground-based data acquired on the last 4 acquisition dates (from 62 to 99 DAP). Only indices with the best prediction performance for both datasets are presented (the complete results are shown in the Table A1).
Traits | UAV Measurements | ||
---|---|---|---|
Vegetation Index | R2 | RMSE 1 | |
Leaf chlorophyll | MTCI | 0.91 | 2.31 |
Leaf area index | TCARI | 0.56 | 0.68 |
Canopy chlorophyll | CIg | 0.81 | 0.201 |
Ground cover | NDVI | 0.90 | 0.037 |
Ground-Based Measurements | |||
Leaf chlorophyll | TCARI/OSAVIre | 0.78 | 3.52 |
Leaf area index | OSAVI | 0.58 | 0.63 |
Canopy chlorophyll | CIg | 0.83 | 0.194 |
Ground cover | OSAVI | 0.92 | 0.033 |
1 RMSE values with same units of corresponding crop properties, i.e., µg·cm−2, m2·m−2 and g·m−2 for leaf chlorophyll, leaf area index and canopy chlorophyll, respectively.