Reconstructing flash flood magnitudes using 'Structure-from-Motion': A rapid assessment tool

- Structure-from-Motion permits rapid field survey as part of post-flood analysis.
- High water marks can be extracted from the resulting model directly.
- SfM coupled with 2D hydraulic modelling offers estimates of peak flood magnitude.

SummaryAccurate records of flash flood magnitudes are required to inform flood forecasting and planning. However, whilst a distributed flood survey is desirable to capture spatial heterogeneity in peak water surface elevation, the field time required for a distributed survey often limits the spatial coverage of such reconstructions. For the first time, we demonstrate the application of Structure-from-Motion (SfM) with Multi-View Stereo (MVS) to reconstruct the magnitude of a flash flood. This approach required only standard digital photographs and ground control points, took only ∼30 min in the field, and can be embedded within existing protocols easily. We validated the method against a conventional dGPS survey in three stages: (i) comparison of topographic data revealed that SfM was accurate to within 0.1 m; (ii) high water marks extracted from the SfM model were within 0.25 m of those surveyed in the field with no consistent over or under-estimate; (iii) peak discharge reconstructed from a two-dimensional hydraulic model was within the range of more conventional estimates. With low uncertainty in our terrain model and our reconstructed flood water surface, we highlight the added value of the SfM approach for incorporating reach scale spatial variability into hydraulic reconstructions.