I wish to add my thank you to Mark Fonstad’s comments. It is a great pleasure to have our work on Self-Organized Criticality in Riverbank Systems recognized in this way.
This award is particularly meaningful to me because the AAG is my professional home, and within that home, the Geomorphology Specialty Group is my family. That sense of family is particularly profound at this moment. The first professional talk I ever gave was given under the auspices of the Geomorphology Specialty Group. The specialty group award that will be given after this one, the Melvin G. Marcus Distinguished Career Award, is named after my father; a fact that still brings great pleasure, joy and an occasional tear to me and my family.
But the family connections go deeper today. As some of you know, it was with great sorrow that we witnessed the loss of Mark’s mother, Karen Wynn Fonstad, to cancer just three weeks ago. Karen was internationally known as a cartographer, but not for the types of maps we classically produce in our research environments. Rather, she was the author of atlases of the imagination, including the Atlas of Middle Earth, The Atlas of Pern, the Atlas of the Land, and many others. A small joy in the sorrowful time of losing Karen is that the Awards Committee notified us of this award well over a month ago. Mark’s mother and father, a physical geographer like Mark, thus knew of this award prior her death. And this is an award that brought Karen real pleasure, not only because Mark is her son, but because she contributed the cartography to the article. As a contributor she thus shares in this award. As for me, the fact that Karen contributed to this article now means as much to me as the award itself. My thanks to the specialty group are thus doubly felt; no one could know just how timely, poignant and special this award would be. Thank you.
There are also two members of the Geomorphology Specialty Group who unknowingly contributed much of the methodological and intellectual foundation for what went into this article. Mike Slattery had the integrity to stand up and give a talk at AAGs some years in which he talked about trying to link channel change to upstream disturbance. He used the classic approach of documenting channel change with cross sections. He reported with honesty and clarity that despite his teams’ extensive field work and analysis, they were unable to demonstrate significant links between morphologic change and upstream impacts – even though he knew that linkages must exist. It was this talk that led Mark and me to develop the mapping scheme for our research, in which we used the “100 m sprint technique” where we mapped a large suite of variables every 100 m for 180 km of stream bank. The moral to this tale – reporting negative results is important. Thank you, Michael.
Mark and I also owe a deep debt to Jonathan Phillips, whose unswerving commitment to evaluating the potential for nonlinear behavior in geomorphic systems has profoundly influenced our thinking over the years. Jonathan had led by example, always providing critical citations from outside geography that demonstrate proof of concept, strong empirical evidence, and qualitative and quantitative models of the system. It was Jonathan’s thinking that opened our eyes to the possibility of SOC behavior; it was his example that helped us structure our article.
Finally, I want to take a moment to philosophize about research directions in geomorphology. Our present research approaches for studying contemporary geomorphic systems are dominated by mechanistic, process-based approaches - as they should be. But at the heart of these approaches is a giant assumption that we sometimes talk about, yet rarely explicitly evaluate – the assumption that we can scale up from local processes to system-wide prediction.
But what if a system is in a self-organized critical state? If this is the case, then the system-wide distribution of magnitude and frequency of events are knowable, but this system-wide knowledge cannot be used to predict behavior at a given point, nor can understanding at one point (or even a lot of points) be used to explain system-wide distributions. We can’t scale up, or down, in an SOC system. In an SOC system, one can explain individual changes after the fact, but can never foresee them.
The system, at least in this sense, is unknowable.
And who among us hasn’t experienced this sense of the unknowable at an intuitive level? Sitting by the stream, year after year, I always have felt on the cusp of revelation, as though the stream was about to speak to me in its secret language, reveal its heart in a way that would let me finally, at long last, look at its parts and explain the whole! But then, as though to spurn me, the stream will shift in a wild and unexpected way – the unexpected log carried from upstream that chokes the channel, the sediment slug cut loose from a debris flow, the abrupt unraveling of an armored bed, the bank collapse aided by a rodent hole - all work to thwart me, year in, year out. I could always explain the change after the fact, but never could have foreseen it.
As we look at rivers, we should therefore think of testing the often untested and unstated assumption - that process models can be scaled up from local physics to predict basin scale behavior. Otherwise, we risk that trap that Roger Coates warned us of in his 1713 Preface to Sir Isaac Newton’s Principia Mathematica:
Those who assume hypotheses as first principles of their speculations… may indeed form an ingenious romance, but a romance it will still be.
Do I even dare point to all the situations where we (including myself) form models based on scaling up, only to create a “romance?” Insofar as we wish to use process models to predict behaviors or identify cause and effect, we should ask this question with regard to: distributed watershed models, sediment transport equations, even Mark’s and my favorite – cellular automata, to name just a few examples.
Rather than focusing on expanding evermore the application of mechanistic models to basin scales, which is our discipline’s trajectory of research endeavor, Mark and I argue that more research needs to focus on first principles of system behavior. If we wish to avoid pursuing illusory outcomes, we should ask more often: “Is the system chaotic, at the edge of chaos, or predictable?” or “Can we scale processes up or down in this system?” Might there be some geomorphic equivalent to the Heisenberg Principle, which by demonstrating the limits of knowledge, radically improved our ability to understand the nature of matter and energy?
Mind you, neither Mark nor I saying all systems are SOC or non-linear. In fact, rather surprisingly for people who have just won an award for research into SOC, we are not even certain that SOC exists as a distinct organizational framework in nature. It sometimes seems to us that SOC is an elegant conjecture built up to explain a consistent pattern – a pattern that might be explained by multiple other organizing theories. But most scientists working with SOC are willing to admit that SOC is not proven and is based on a series of unclear assumptions. Can we say the same thing when we scale up with so-called “deterministic” models to explain entire river systems?
To those of you familiar with literature on nonlinear systems, there is nothing new in what we are saying. But given the emphasis we continue to see in much geomorphic research, we believe it is worth reiterating.
Oscar Wilde had it right when he said “The true mystery of the world is the visible, not the invisible.” In an SOC system, our inability to predict is NOT because a process or parameter is invisible to us; it’s NOT because we need to measure more or model better. Rather, the system organizes itself, right before our very eyes, in ways that are fluid, mysterious and beyond our ability to know. Is this not a beautiful thing; to know its form, but to find it forever perplexing? It is this tantalizing mystery, the desire to grab and hold this shifting ghost-like creature, almost seen, but never entirely perceived, that draws us again and again to rivers. It is why, I believe, most of us who are river scientists are - in the very truest sense of the word – “haunted” by rivers. If we not only acknowledge this mystery, but embrace it, try to define what we can NOT know as well what we can, then our fascination, our appreciation, and our understanding of rivers will only grow.
Back to Marcus home page