Volume estimation for high-locality fragmented rockfall using UAV-based photogrammetry

Empirical-statistical and field measurement schemes for high-locality fragmental rockfall volume estimation are challenging to obtain an accurate and reliable result. The flexible and adaptive statistical method using remote sensing technology may improve the quality of rockfall volume estimation which is important for hazard assessment. In this study, a hybrid methodology for the volume estimation in fragmental rockfall events is presented. The image recognition techniques combined with an Unmanned Aerial Vehicle (UAV) are used to estimate the block sizes in the deposit area. Compared to field-measured values, the relative errors are less than 6% indicating the feasibility of the proposed method in a rockfall block size estimation. Therefore, the fragmental rockfall volume can be determined based on the Rockfall Block Size Distribution (RBSD). The RBSD of fragmental rockfall can be well-fitted by a power-law distribution ( $$\:y=0.01{V}_{0}^{-1.14}$$ ). Then, the estimated volume is compared to the result from pre- and post-failure changes in the surface elevation by the Digital Surface Model (DSM). The mean ratio is up to 82.26% based on the depletion volume, and 90.65% on the deposition volume. The estimation accuracy is better than the ratio of 57% to empirical formulas for the rockfall volume estimation. Even though there are still uncertainties in the volume estimation, the results show that the proposed method may be helpful for such kind of hazard assessment and mitigation.