Table 3 Summaries of classification accuracy
| Â | Area | Without NIR-derived variables | With NIR-derived variables |
|---|---|---|---|
Median (IQR) | Median (IQR) | ||
Overall accuracy | Interpolated | 0.912 (0.878–0.916) | 0.907 (0.902–0.915) |
|  | Extrapolated | 0.761 (0.754–0.770) | 0.790 (0.753–0.808) |
Surface water accuracy | Interpolated | 0.983 (0.980–0.987) | 0.983 (0.979–0.986) |
Extrapolated | 0.942 (0.939–0.945) | 0.935 (0.932–0.938) | |
Overall kappa | Interpolated | 0.903 (0.894–0.908) | 0.898 (0.892–0.907) |
Extrapolated | 0.738 (0.731–0.748) | 0.770 (0.730–0.790) | |
Surface water kappa | Interpolated | 0.960 (0.953–0.970) | 0.960 (0.951–0.967) |
Extrapolated | 0.871 (0.866–0.878) | 0.856 (0.848–0.863) | |
Overall quantity disagreement | Interpolated | 21 (17–24) | 22 (19–26) |
|  | Extrapolated | 70 (65–74) | 63 (59–66) |
Surface water quantity disagreement | Interpolated | 3 (2–5) | 7 (3–11) |
|  | Extrapolated | 3 (2–6) | 24 (9–32) |
Overall allocation disagreement | Interpolated | 43 (38–46) | 42 (37–48) |
Extrapolated | 95 (91–101) | 81 (69–107) | |
Surface water allocation disagreement | Interpolated | 8 (6–10) | 4 (2–8) |
Extrapolated | 36 (34–38) | 18 (12–36) | |
% surface water | All | 18.3 (17.3–20.9) | 16.6 (15.7–22.0) |