Quantifying the contributions of precipitation, topography and human activity and their coupling to the development of permanent gully

• Linear, areal, and volumetric gully erosion rates were quantified during 2013–2021. • Ridge orientation, road and shelterbelt had different effects on gully erosion rate. • Horizontal ridge tillage can promote headcut retreat under heavy precipitation. • Tillage downslope on both sides of gully is recommended to limit gully expansion. • The coupling of topography and human activity mainly contributes to gully erosion. Gully erosion poses greater threat to ecological and food security globally. The gully development rate and the contributions of primary influencing factors have not been thoroughly understood, primarily due to the complicated environmental conditions characterized by climate, soil, topography, and human activities. Therefore, this study was conducted to explore the gully development and contributions of its driving factors from 2013 to 2021 based on historical Google Earth images and field investigation in a typical rolling hilly watershed of northeast China. The climatic factors reflected by precipitation, topographical factors related to gully, and anthropological factors reflected by tillage ridge orientation and linear anthropological factors (LAF) including roads and shelterbelts were obtained. The tillage effect index (TEI), LAF effect index (LEI) and the distance from gully to LAF (D L ) were calculated to analyze the influence of anthropological factors on gully development. The results showed that the linear, areal and volumetric gully development rates (R L , R A and R V ) showed great variabilities in different periods with the average of 5.69 m year −1 , 137.37 m 2 year −1 and 428.54 m 3 year −1 , respectively. The accumulated R L , R A and R V increased exponentially or linearly with accumulated precipitation amount (P), erosive rainfall duration (RD), erosive rainfall (ER), maximum 30-min rainfall intensity (I30) and rainfall erosivity (RE) increased (P < 0.001). The slope of the gully head (HS) was identified as the primary topographic factor influencing R L . Both R A and R V showed positive power relationships with the drainage area of the gully (CA) and the upstream gully head (HA). The human activity reflected by tillage ridge orientation, unpaved road and shelterbelts had different effects on R L and R A . The increased orientation angle between gully and tillage ridge (ROA) mitigated headcut retreat for normal precipitation year but it would aggravate headcut retreat in the periods of continuous heavy rainfall. Gully bank with the ROA of 30−90° was more prone to areal development. The TEI, LEI and D L had significantly positive linear, power or exponential relations with gully development rates. The PLSR and VPA analysis indicated that topography, the interaction between topography and human activity, and the coupled effects of precipitation, topography and human activity were the main contributor of R L , R A and R V , respectively, and accounted for 34 %, 50 %, and 40 % of total variability of R L , R A and R A , respectively. This study can provide guidance for the control of gully erosion and the optimization of land use structure in the gully catchment.