CSDMS 2016 annual meeting poster AnthonyLongjas
An entropy based quantification of delta channel network complexity
Alejandro Tejedor, St. Anthony Falls Laboratory, University of Minnesota Minneapolis Minnesota, United States. alej.tejedor@gmail.com
Efi Foufoula-Georgiou, St. Anthony Falls Laboratory, University of Minnesota Minneapolis Minnesota, United States. efi@umn.edu
Abstract:
River deltas across the world exhibit an astonishing variety of morphologies in response to different forcings (e.g., river, tides and waves), sediment composition, incoming flow variability, sea level rise, etc. Understanding and quantifying the patterns imprinted on the landform would enable us to infer processes from observed imagery. Galloway [1975] introduced a qualitative diagram to classify deltas, showing how the balance of upstream (fluvial) and downstream (waves and tides) forcings dictates the delta form, depicted most distinctively in the coastline morphology. Recently, we presented a rigorous framework [Tejedor et al., 2015a,b] based on spectral graph theory to study delta channel networks, enabling us to extract important structural and dynamics-related information of river deltas. Using that information, we are able to introduce a suite of metrics to quantify channel network complexity, including entropic-based metrics measuring the complexity in terms of the uncertainty in the splitting and rejoining paths and fluxes, enhancing the comparison of deltas and process from form inference. Finally using the above mentioned framework, we are able to construct vulnerability maps that depict the relative change of sediment and water delivery to the shoreline outlets in response to possible perturbations in hundreds of upstream links. We show that an inverse relationship exists between entropy and vulnerability, reinforcing the idea that entropy is a surrogate of the capacity of the system to undergo changes.
* Please acknowledge the original contributors when you are using this material. If there are any copyright issues, please let us know and we will respond as soon as possible.
