부산대학교 도서관

A technique is presented and tested which enables land surface roughness on the vertical scale of a few centimetres to be estimated from an aircraft-mounted scanning laser altimeter, or LiDAR (Light Direction And Ranging) system. While the systematic uncertainty in measuring land elevation using a LiDAR system is around 20 cm, we demonstrate that sequential laser pulse returns are highly correlated in height, and after detrending to remove the effects of topographic slope, show variation that can enable different physical soil treatments to be distinguished. Test sites were created on Sonning Farm in Reading, UK, with treatments including rolling flat, ploughing and potato-ridging. The standard deviation of detrended returns from 26 LiDAR acquisitions show a clear variation between the different treatments which would enable surface roughness to be estimated without fieldwork. This provides a useful variable to models of overland water flow which will strongly influence sediment erosion, transport and deposition and flood extent.