Monitoring rill and ephemeral gully erosion by Structure from Motion (SfM) technique

Abstract

The first part of this investigation was aimed to test the reliability of the measurements of morphological and hydraulic variables by the Structure from Motion (SfM) technique for a single rill channel. These measurements were compared with those carried out on the rill gypsum cast. Also, for two closed earth channels having a different sinuosity, the ground measurement of some cross sections by the three‐dimensional (3D) digital terrain model and DEM, derived from SfM, and a profilometer was carried out. Moreover, the real volume of the two channels was compared with the volumes measured by the three methods. The results pointed out the high reliability of the SfM technique and that the 3D model can be used as a benchmark. Then, two field applications at the Sparacia experimental area (Sicily, Southern Italy) were reported. Terrestrial and low‐altitude aerial surveys were carried out by consumer grade camera and low‐cost unmanned aerial vehicle (UAV) to measure rill erosion on two plots with steepness of 22% and 26%. Furthermore, the 3D model and DEM of a plot were built using the available aerial photographs (166) and adding 40 terrestrial photographs. The comparison among the three surveys (terrestrial, UAV, and UAV + terrestrial) was developed using two morphometric parameters of the rill network (drainage density and drainage frequency). The morphometric parameters were higher for the terrestrial than for UAV and UAV + terrestrial surveys. For both plots, SfM always provided reliable soil loss measurements, which were affected by errors ranging from −8% to 13% as compared to the reference weight of the sediment stored in the tanks located downstream of the plots. The three ground measurement methods (3D model, DEM and profilometer) were also applied for surveying an ephemeral gully (EG) channel. The comparison between the values of morphological and hydraulic variables surveyed at 24 cross sections by both DEM and profilometer showed that the estimate error was generally less than ±10%. The EG measurements carried out by the three methods supported the applicability of an empirical relationship between EG length and its eroded volume. Finally, it was demonstrated that the effect of the distance interval on the EG volume measurement by 3D model is negligible for the investigated EG.