Landslide erosion coupled to tectonics and river incision

The steep topography of mountain landscapes arises from interactions among tectonic rock uplift, valley incision and landslide erosion on hillslopes. Hillslopes in rapidly uplifting landscapes are thought to respond to river incision into bedrock by steepening to a maximum stable or ‘threshold’ angle1–3. Landslide erosion rates are predicted to increase nonlinearly as hillslope angles approach the threshold angle1–7. However, the key tenet of this emerging threshold hillslope model of landscape evolution—the coupled response of landslide erosion to tectonic and fluvial forcing—remains untested. Here we quantify landslide erosion rates in the eastern Himalaya, based on mapping more than 15,000 landslides on satellite images. We show that landslide erosion rates are significantly correlated with exhumation rates and stream power and that small increases in mean hillslope angles beyond 30 translate into large and significant increases in landslide erosion. Extensive landsliding in response to a large outburst flood indicates that lateral river erosion is a key driver of landslide erosion on threshold hillslopes. Our results confirm the existence of threshold hillslopes and demonstrate that an increase in landslide erosion rates, rather than steepened hillslope angles, is the primary mechanism by which steep uplands respond to and balance rapid rates of rock uplift and bedrock river incision in tectonically active mountain belts. The threshold hillslope paradigm is rooted in the observation that hillslope angles throughout mountainous landscapes tend to be symmetrically distributed about a mean value with a mode comparable to the friction angle of granular material1,2,8–10. In contrast to how hillslope angles and erosion rates increase linearly to keep pace with rock uplift in landscapes with low to moderate tectonic forcing3,5–7,11, at high uplift rates hillslope angles are thought to be limited by material strength12, so hillslopes will approach the threshold angle and erosion rates will increase nonlinearly such that the relationship between erosion rates and slope angles approaches asymptotic. Vertical river incision into bedrock is thought to over-steepen hillslopes with gradients near the threshold angle, increasing relief until gravitational stress exceeds material strength and bedrock landsliding occurs1. Hence landscapes with hillslope gradients near the threshold angle are thought to respond to increases in uplift-driven river incision by increasing landslide erosion rates, rather than by steepening1,3,5. Implicit in the threshold hillslope model are the assumptions that landslide erosion rates spatially track rates of river incision and that landslide erosion rates increase nonlinearly as hillslope gradients approach the threshold angle. In steady-state landscapes, the threshold hillslope model also predicts that landslide erosion rates are spatially coupled with exhumation and rock uplift rates. Strong