Degree
Doctor of Philosophy (PhD)
Department
Geology & Geophysics
Document Type
Dissertation
Abstract
Landscape evolution models (LEMs) and detrital zircon provenance are widely used to investigate Earth surface processes, yet both face challenges related to quantifying model behavior and interpreting sedimentary signals under realistic, heterogeneous conditions. This thesis addresses these challenges through analysis and case studies that develop and apply new approaches for analyzing landscape dynamics and assessing uncertainties in sediment provenance.
Chapter 1 introduces a quantitative framework for evaluating LEM outputs using Shannon Entropy, Moran’s I, and Geary’s C. Applied to simulations with steady uplift, periodic uplift, spatially variable uplift, and a timeseries of DEMs, these metrics capture spatial organization and information content within and across model runs. This approach enables systematic detection of divergence, variability, and spatial coherence in model behavior, offering a flexible toolset for improving comparisons among simulations and for analyzing any matrix-based topographic dataset.
Chapter 2 examines key assumptions underlying detrital zircon provenance analysis, specifically the uniformity of erosion rates, zircon fertility, and zircon grain-size distributions within source areas. Using suites of statistical simulations, the influence of parameter variations on provenance age distributions was evaluated, along with the extent to which models weighted by area, zircon mass, or zircon abundance diverge. Results show that population complexity, parameter variation, and sample size jointly control the distinctiveness of provenance signals, with larger sample sizes reducing variance and parameter differences generally producing subtle effects when realistic source populations are used.
Chapter 3 links surface processes directly to detrital zircon signals by integrating landscape evolution modeling with zircon transport and sampling. Using Landlab and coupled erosion–sediment transport components, the influence of variable rock erodibility, layered stratigraphy, zircon fertility, and zircon grain size on the time-dependent provenance record is evaluated. The experiments demonstrate that surface processes imprint measurable and sometimes systematic biases on detrital zircon distributions, and that zircon-specific properties can either amplify or moderate these effects.
These studies examine quantitative methods for analyzing model outputs and investigate how landscape dynamics and zircon characteristics shape sedimentary signals, highlighting the importance of moving beyond simplifying assumptions and embracing spatial variability to better understand how Earth’s surface processes are recorded in both models and provenance.
Date
1-1-2026
Recommended Citation
Carmello Grom, Vivian, "Quantifying Uncertainty In Landscape Evolution And Provenance Analysis: Insights From Statistical And Numerical Modelling" (2026). LSU Doctoral Dissertations. 6974.
https://repository.lsu.edu/gradschool_dissertations/6974
Committee Chair
Adam Forte