Layer-Scale Spectral–Sediment Relationships and Calibrated Uncertainty from Underwater Hyperspectral Observations During an Upper-Yangtze Flood Event

Within a single flood event, this study evaluated the empirical relationship between underwater hyperspectral observations and layer-scale suspended sediment concentration (SSC) and calibrated its predictive uncertainty. The dataset comprised 101 paired observations from 25 complete four-position verticals (n = 100) plus one isolated record. Sample-level leave-one-out cross-validation (LOOCV) provided the primary estimate of event-internal layer-level interpolation, while leave-one-vertical-out validation (LOVO) assessed sensitivity to an unseen vertical. Under LOOCV, random-forest models reached R2 ≈ 0.78–0.79 and RMSE ≈ 0.087 kg/m3; on the complete-vertical subset, spectra-only RF_log1p decreased to R2 = 0.595 and RMSE = 0.120 kg/m3 under LOVO. Discharge and interpolated stage added event-specific context (Q/stage-only LOVO R2 = 0.699), but cross-date transfer remained poor. Raw quantile-regression-forest intervals under-covered the nominal 90% level; conformal calibration restored coverage at the cost of wider intervals (grouped PICP90 0.721 → 0.927; MPIW90 0.234 → 0.617 kg/m3). The four positions showed a weak within-vertical effect (Friedman χ2 = 11.93, p = 0.0076; Kendall’s W = 0.159), with no pairwise contrast significant after Holm correction. The results support calibrated event-internal interpolation at sampled layers, but do not resolve a monotonic vertical profile or support transfer across dates, stages, stations, or events without recalibration.