Photo: Rating curve training course at Canadian River near Konawa , Oklahoma. The author is the genius wearing jeans in 100 F heat. Monty Porter is in the centre wearing the red cap, blue shirt and tan shorts.
There may never be busloads of tourists traveling the world to visit places with interesting rating curves but if there were then Oklahoma would be a tourist mecca.
I was asked by the Oklahoma Water Resources Board to provide some on-site rating curve training and I was initially a bit lukewarm to Oklahoma as a destination. That changed when Monty Porter explained the diversity of channel types that they deal with. I was sold when he described their sand channel streams.
Sand channels are to hydrographers what an ironman competition is to an athlete.
It may take years of preparation to prepare for the challenge. It takes excellent observational skill and aptitude; a solid understanding of the underlying principles of hydraulics and fluvial geomorphology; and the ability to think of the rating not simply as a one dimensional line but rather as a two dimensional surface where the extra dimension is time.
Rating curves are really the only game in town for sand channel streams.
In many other types of streams if there are problems defining the rating then it is possible to put in additional instrumentation to support a more complex model such as stage-slope or index velocity. Simply monitoring stage in a constantly moving sand channel is challenging enough.
The approach we took for this course was to start with a full day in the field. It is only in this context that discussion about hydraulics, sediment transport, fluvial geomorphology, hydrology and riparian biology really make sense. With nothing more than our powers of observation, a stadia rod and the current water level we were able to estimate the values to use for the offset and the exponent for the rating curve. We were able to identify breakpoint elevations and estimate the effect of the break point on the value of the offset and exponent. We were able to develop a first order approximation of the effect of vegetation on the relation, and in particular the effect of vegetation that had matured over the course of a three year drought and was now armoring the tops of the sand bars. By discussing the wave types previously observed at different water levels we were able to develop an crude understanding of the evolution of bed roughness as a function of flow: plane bed; rippled; dune; and anti-dune.
Bringing these observations and deductions back into the classroom provided an excellent starting point for rating curve development. We were readily able to fit a curve to our rating measurements starting from the parameters identified in the field with minor adjustments. The next challenge was accommodating the dynamic component of the rating, which we could predict based on our deductions and observations of the sediment size, channel bed and riparian vegetation. What we found was that for lack of detailed observations (e.g. wave type amplitude and frequency) or photographs attached to measurements that we needed to rely on interpreting our conceptual model in the context of the stage hydrograph to model the deviations from the base rating. This was enough to provide satisfying results but we all agreed that we could have done a better job if we better descriptive observations and photographs attached to the rating measurements.
You may not need to travel the world to find challenging rating curve problems.
There always seems to be too much hydraulic complexity for the distribution of rating measurements to explain in a simple empirical model. Sand channel streams are an excellent example of how you can explain a complex underlying relation with a relatively simple model. The shape of the shift corrections is determined by the rating measurements (classified by the hydraulic condition at the time of the measurement) but decisions about how to pro-rate the shifts into each other is informed by inspection of the stage hydrograph in the context of knowledge (observations and theory) about bed movement (armoring, grain size distribution, depth, kinetic energy, density and viscosity etc.).
A reliable rating curve is one that is credible, defensible, and minimizes re-work. This paper outlines 5 modern best practices used by highly effective hydrographers.