Environmental Controls on the Lateral Distribution of Stromatolites

Abstract

Stromatolites are mounded sedimentary rocks formed from layers of cyanobacteria. They exist in carbonate units across the world and are the most abundant fossil record of life in the Precambrian era. Understanding what controls the area, connectivity, and depth of laterally adjacent sedimentary environments is necessary in order to know what the extent and continuity of reservoirs is in the rock record. However, little is known about how the local environment controls the reservoir characteristics of stromatolite units. Using a combination of drone imagery, computer vision techniques, and field data from over 1100 stromatolites, I present the first quantitative analysis of the morphology and connectedness of modern-day living stromatolites in Shark Bay, Western Australia. I create a detailed bathymetric map of the region and develop machine-learning algorithms that isolate individual stromatolites as well as distinct patterns within them. With these data, I test the hypothesis that subtle changes in seafloor elevation and gradient are the main control on the amalgamation and coalescence of stromatolites to make connected reservoirs, and that their morphology encodes information about the environment in which stromatolites grow.