Influence of precipitation change and topography characteristics on the development of farmland gully in the black soil region of northeast China

Abstract

Gully erosion in cultivated land has seriously damaged land resources and reduced crop production. The long-term changes of gully erosion have not been well understood because of limited continuous observed data on the gully development process, which further causes difficulty in establishing universal models of gully erosion rates. Hillslope gullies and valley floor gullies were continuously monitored by RTK GPS and Unmanned Aerial Vehicle (UAV) from 2002 to 2021 in the Heshan farm of Songnen black soil region of northern China. UAV-generated DEMs, historical remote sensing images, and daily precipitation data were combined to analyze the influence of precipitation changes and topographic differences on the development process of gully erosion. The results showed that gully erosion had strong variability between years due to the difference in annual precipitation. The volumetric increment in high precipitation years was 6.5 times and 16.5 times that in normal precipitation years and low precipitation years, respectively, and the variability of valley floor gullies was greater than that of hillslope gullies. The annual average headcut retreat of valley floor gullies and hillslope gullies was 13.1 m y-1 and 4.2 m y-1, respectively, but the volumetric soil loss was mainly contributed by gully bottom deepening and gully sidewall collapsing. The ratio of volumetric soil loss from the gully head to total volumetric soil loss was the smallest in high precipitation years, followed by normal precipitation years and low precipitation years because more runoff in high precipitation years enhanced erosion kinetic energy, which accelerated gully bottom deepening and gully sidewall collapsing. Topography and rainfall were the important influencing factors for the continuous gully development. The maximum value of 3-day accumulative rainfall best explained valley floor gully erosion, and the accumulative value of erosive rainfall best accounted for hillslope gully erosion. A statistical model for estimating volumetric soil loss rate (△V) was developed that included as independent variables drainage area (A) and rainfall characteristic index (RI): △V = 1.571*10-7*（A*RI）1.983 (R2 = 0.98).