Super-resolution mapping of gravel bed surface roughness using terrestrial laser scanning and airborne laser scanning

The aim of this study is to derive the scaling behavior of gravel-bed roughness of terrestrial laser scanning (TLS) from airborne laser scanning (ALS). We use the fractal dimension, which can be estimated from the slope of log-log variogram of laser scanning data, to represent gravel-bed roughness. The study area is located near the confluence of the Nan-Shih River and Pei-Shih River, northern Taiwan. Two gravel-bed sites with the extent of 6m × 6m were acquired by FARO photon 80, and then in-situ digital surface models (DSMs) were generated with 1 cm × 1cm resolution. ALS survey was conducted by an Optech ALTM 3070 at the density of 247 points/m². The point cloud data were collected by ALS, which showed smoothed surface of TLS counterparts. In this study, we used the deconvolution procedure to derive the variograms of TLS-derived DSM from the variograms of ALS data. Hence, we could establish the link between the gravel-bed roughness of TLS-derived DSM and ALS data. The results will lead a better understanding of gravel-bed roughness using ALS data for the large-scale areas.