Mapping of Rill Erosion of Arable Soils Based on Unmanned Aerial Vehicles Survey

• Published in: Eurasian soil science 2018-04, Vol.51 (4), p.479-484
• Main Authors: Kashtanov, A. N., Vernyuk, Yu. I., Savin, I. Yu, Shchepot’ev, V. V., Dokukin, P. A., Sharychev, D. V., Li, K. A.
• Format: Article
• Language: English
• Subjects: Aerial surveying; Agricultural land; Arable land; Cameras; Conservation of Soils; Data processing; Degradation; Detection; Drone aircraft; Earth and Environmental Science; Earth Sciences; Erosion; Forest soils; Geographic information systems; Geographical information systems; Geotechnical Engineering & Applied Earth Sciences; Height; Identification; Mapping; Polls & surveys; Procedures; Rehabilitation; Remote monitoring; Remote sensing; Rill erosion; Rills; Satellite navigation systems; Seasons; Snowmelt; Soil; Soil erosion; Soil mapping; Soil surfaces; Unmanned aerial vehicles; Wheat; Winter wheat
• ISSN: 1064-2293; 1556-195X
• DOI: 10.1134/S1064229318040051

Possibilities of using data obtained from unmanned aerial vehicles for detection and mapping of rill erosion on arable lands are analyzed. Identification and mapping of rill erosion was performed on a key plot with a predominance of arable gray forest soils (Greyzemic Phaeozems) under winter wheat in Tula oblast. This plot was surveyed from different heights and in different periods to determine the reliability of identification of rill erosion on the basis of automated procedures in a GIS. It was found that, despite changes in the pattern of rills during the warm season, only one survey during this season is sufficient for adequate assessment of the area of eroded soils. According to our data, the most reliable identification of rill erosion is based on the aerial survey from the height of 50 m above the soil surface. When the height of the flight is more than 200 m, erosional rills virtually escape identification. The efficiency of identification depends on the type of crops, their status, and time of the survey. The surveys of bare soil surface in periods with maximum possible interval from the previous rain or snowmelt season are most efficient. The results of our study can be used in the systems of remote sensing monitoring of erosional processes on arable fields. Application of multiand hyperspectral cameras can improve the efficiency of monitoring.