Understanding Emergent Phenomena in Barchan Swarms Using an Agent-Based Model

Much of the behaviour of isolated barchans - for instance the existence of a minimal size and the size-dependence of migration rates - is well understood and can be predicted using simplistic models of sand transport.  However, in most instances, barchans do not occur as solitary bedforms but appear in large populations known as swarms.  One can find vast examples of these systems extending for many kilometers and containing tens of thousands of dunes on both Earth and Mars.  Within these swarms, the individual dunes interact through manipulation of the sand flux field which occurs as upwind dunes absorb incoming flux across their entire width and emit flux only through their horns.  Furthermore, the different migration rates of the bedforms lead to collisions which result in the redistribution of mass between the dunes and can also lead to the destruction and creation of barchans. 

  The interactions between barchans in a swarm lead to many emergent phenomena which our knowledge of the isolated bedforms cannot explain.  Several studies have sought to understand, perhaps the most well-documented of these properties, size selection.  However, there has been less attention given to the role played by interactions in governing the spatial structuring of swarms.  It is known, for instance, that barchans tend to align with the horns of their upwind neighbours, this can lead to the formation of striking echelon patterns.  Other reported emergent spatial phenomena include homogeneity of inter-dune spacing and periodicity in spatial correlation functions.  In this presentation we will describe a novel agent-based model we have constructed and discuss the insights it can provide into the nature of the different emergent properties within barchan swarms.  We will compare the results of the model to observations of real-world swarms on Mars and Earth.