Transition Analysis for River Morphology

Kernel_Smoothed

Water is most decidedly a crucial resource for the Ucross ranch. Fueled to a large extent by snowmelt from the Bighorns, Piney and Clear Creeks are the major streams that run through the property. Understanding how these two streams operate helps inform management when making decisions on where to, for example, stabilize a bank so that the water doesn’t erode pasture areas for grazing. But what’s the best way to understand how these streams operate and how they change over time?

Modeling of all types, shapes, and sizes has become one of the most widely used strategies in contemporary science for understanding how natural systems operate. Accordingly, the UHPSI team is experimenting with a new type of “transition matrix modeling” for river morphology (i.e. how the river structure changes over time). This method of modeling takes a probabilistic approach to understanding the river equilibrium. Instead of attempting to create a complex mathematical function that can predict what might happen in the future, the team analyzed what has happened to the river in the past ≈60 years according to available aerial and satellite imagery then uses this information to assign a probability on how the river might change in the future. The more the river has demonstrated change in its structure within the past ≈60 years, the higher the probability that it might change in the future. Importantly, these models allow for the incorporation of new data (in the form of technologically advanced satellite and aerial imagery) within its structure, meaning that the model dynamically changes as the system itself changes. This type of model could be used to help form a metric for helping scientists and managers understand the “level of dynamism” in a river system, which is to say it gives them an actual scale on which to understand the likelihood of a river changing structure over time. If a manager is interested in finding the areas that are the “most dynamic” (and therefore potentially in need of stabilization in some way), he/she could view a map like the one above, which shows the confluence of Piney and Clear Creeks on the southern tip of the Ucross Ranch. The dark blue color indicates an area that exhibited the “most change” across the timeframe we analyzed; the yellow color indicates areas that did not change substantially; blueish-yellow indicates areas that changed to some degree.

For more information on this particular modeling method, email devin.routh@yale.edu.