Quantitative Study On Rill Erosion In The Northeast Slope Farmland Based On The UAV Micro-Topography Detection

Rill erosion is a common form of soil erosion in the black soil area of Northeast China,especially in the snowmelt period,the concentrated runoff often aggravates rill erosion and leads to more soil loss.However,the most of the actual soil dislocation on eroding slopes cannot reach the field outlet or catchment gauges and thus cannot be measured in time.In this study,we conducted a series of run-on experiments on a ridged and sloped cropland in Northeast black soil area to simulate snowmelt runoff induced rill erosion.Three different UAV flight schemes were carried out to generate digital surface model(DSM)of different resolutions and compare the calculated rill erosion volumes via grid subtraction.Moreover,fine sand was refilled into selected rill segments and the actual volumes were recorded to verify the accuracy of the sand volume calculated by UAV images.Finally,to identify the limitation of traditional rill erosion methods we compared the calculated rill erosion volumes with that estimated from the ratio of width to depth,and constant cross-section.The major findings are listed as follows.(1)The rill erosion volume can be accurately extracted by a low-flying drone with a tilted camera combined with grid subtraction method.When rill erosion depth > 1 cm,the calculation accuracy of erosion volume was 87.9%under UAV flight scheme 1(10 m height with a tilted camera),which proves the usability of this UAV flight scheme to investigate rill erosion in the situation.The greater the erosion depth is,the higher the calculation accuracy is.(2)To save data acquisition and processing cost,UAV scheme can be simplified when investigate severe rilled area by increasing flight height and using an orthographic camera.Compared with results obtained by the high-precision flight scheme 1,the calculation errors of flight scheme 2(10 m height with an orthographic camera)and flight scheme3(20 m height with a tilted camera)were 9% and 20% respectively,but the data acquisition and processing time reduced by about 50%.(3)The rill erosion volumes obtained by high-precision UAV flight scheme verify the limitations of traditional rill investigation method to simplify rill morphology.For rill segments on upper and middle slope,the run-on water not only causes rill undercutting,but also causes rill bank erosion to a certain extent,thus the traditional ratio of width to depth change had a poor fitting effect with rill erosion volume.Oppositely,for rill segments on lower slope,the rill shaping mode is mostly gully undercutting because of the weak runoff,so the ratio of width to depth change had a good fitting effect with rill erosion volume,with a determination coefficient of 0.76.In addition,compared with the results obtained via grid subtraction,the calculation errors of the traditional cross-section simplified method increase with the decrease of soil erosion depth,and the simplified method is only suitable for rill erosion survey when erosion depth > 1 cm.On the whole,aiming at the problem that rill erosion on long gentle ridged slope land in Northeast China is difficult to be precisely quantified,this study explored the feasibility of UAV micro-topography detection technology to characterize rill erosion precisely and efficiently under the condition of concentrated runoff.The applicability of different UAV flight schemes was analyzed,and the limitations of simplified rill morphology investigation method were also evaluated.The UAV technology overcomes the delay and one-sidedness of the traditional method to estimate the slope soil erosion based on the soil loss amount in the field outlet or catchment gauges,and provides a technical reference for the investigation of rill erosion in the field.