Hypsometry of drainage basins (area±elevation analysis) has generally been used to infer the stage of geomorphic development and to study the influence of varying forcing factors (i.e. tectonics, climate, lithology) on topography. However, the scale dependence of hypsometry has generally been neglected. In order to assess the scale dependence of hypsometry, this study focuses on the sensitivity of hypsometry to different Digital Elevation Model (DEM) resolutions and on the influence of drainage area. Hypsometry inferred from different DEMs is shown to be robust against variations of their resolution. However, hypsometry appears to be dependent on drainage area. We propose that this scale dependence may reflect the varying importance of river and hillslope processes with basin area.

Abstract. The present study investigates correlations between an extensive range of geomorphic properties that can be estimated from a digital elevation model and the uplift rate on geological timescales. The analysis focuses on an area in the Siwalik Hills (central Nepal), where lithology and climate can be considered as uniform. This area undergoes rapid tectonic uplift at rates of up to 15 mm yr -•, which are derived from the geometric pattern of a faultbend model of fold growth. The selected geomorphic properties can be divided in two categories, depending on whether or not the vertical dimension is taken into account. None of the planar properties are significantly correlated to uplift rate, unlike those that include the vertical dimension, such as the mean elevation of basins, hypsometric curve, and hypsometric integral, and relief defined by the amplitude factor of length scaling analysis. Correlation between relief and uplift rate is observed for all length scales of topography shorter than 600 m, which suggests that all orders of the streams are able to adjust to the tectonic signal. Simple mass balance considerations imply that the average elevation is only 10% of surface uplift, suggesting that a dynamic equilibrium has been reached quite rapidly. Using a simple two-process model for erosion, we find that fairly high diffusion coefficients (order of 10 m 2 yr -•) and efficient transport of the material by rivers are required. This unusually high value for mass diffusivity at small length scales may be obtained by either a very efficient linear diffusion or by landsliding. Actually, both processes may be active, which appears likely given the nature of the unconsolidated substratum and the favorable climatic conditions. Local relief in the study area may therefore be used to predict either uplift or denudation, but the prediction is calibrated only for that specific climatic and lithologic conditions and cannot be systematically applied to other contexts.

No abstract