Forward, Inverse and Spectral Analyses of Drainage Networks and Landscapes Reveal Dominance of Regional Tectonics and Mantle Dynamics

Over the last decade, studies have used inverse modeling of drainage networks to reveal quantitative information on tectonic and mantle dynamic forces driving landscape growth, subject to carefully calibrated erosional models. Here, we demonstrate that a range of testable predictions results from landscape evolution models driven by tectonic forcing derived from inversion of river profiles. Predictions of topography, drainage networks, sedimentary flux histories and rock cooling rates from models match observations, despite relaxation of simplifying assumptions which were made by inverse models used to calculate the tectonic forcing. Furthermore, inverse modeling of synthetic rivers generated by these naturalistic landscape evolution models permits accurate recovery of the input uplift history. Wavelet power spectral analysis of observed river profiles reveals self-similarity for long-wavelength features, with an increase in power spectral slope for wavelengths <~ 100 km. This increase in spectral slope means that sharp, i.e. high-amplitude, short-wavelength features, exist only below a certain scale. At longer wavelengths, profiles are dominated by broad signals of tectonics or mantle dynamics which are roughly self-similar. Spectral analysis reveals that drainage networks generated by landscape evolution models subjected to uplift forcing from inversion of observed drainage are self-similar across all scales. Power spectral slopes of predicted rivers are similar to observations when forward models include spatial noise in uplift forcing or erodibility. Sharp signals arise at short wavelengths in response to factors which either operate below inverse model resolution, or do not correlate between tributaries, i.e. from lithologic strength contrasts, human alteration, or self-forming hydraulic shocks. Most topographic power in landscapes resides at long wavelengths, permitting accurate recovery of records of regional uplift. Extrapolating studies of short-wavelength landscape evolution processes across all scales is flawed.