Stage-discharge rating curves define a unique relation between water level and discharge, enabling continuous derivation of streamflow from water level record. This is important because water level (which is relatively easy to monitor) is only locally relevant whereas discharge (which is relatively difficult to measure directly) is the integral of all runoff processes upstream of the gauge. The vast majority of all streamflow data that has ever been produced is a derived result of a rating curve.
In other words, almost everything that we know (or rather that we think we know) about hydrology is a result of rating curves.
The problem with this paradigm is that there has never been global consensus on best practices for the development and management of rating curves. Assuming stable control features, steady flow, and an abundance of gaugings evenly distributed over the full range of stage and time, then regional differences in methods for rating curve development might be interesting but would probably not be very impactful. However, it is rarely the case that any rating curve problem is fully constrained to reveal a unique solution regardless of approach.
Given that ratings are managed in the real world – not some idealized world of fully predictable hydraulic conditions – it’s curious that there has, historically, been very little interest in the dependency of trillions of dollars’ worth of water resource decisions on the truthfulness of rating curves.
It’s almost like a Stockholm syndrome phenomenon.
Hydrologists, engineers, water resource managers, along with the many other users of flow data are held hostage by their dependence on rating curves. They are grateful to get any data at all and loathe to reject data when there really isn’t any realistic alternative to the status quo.
It is therefore up to the hydrometric community to step up to the challenge of setting standards and developing best practices to ensure that the results of rating curve derivation are, at the very least, consistently meaningful.
It’s not the case that all derived discharge results need to have equally low uncertainty.
However, it is the case that if confidence in the derived result is low that there should be some way of communicating that, under the circumstances, an explicit determination of discharge is not possible. Even when the quantity of flow is uncertain the data may be extremely useful for determining the timing, routing, and relative magnitude of peak events and should certainly inform where some change in the monitoring plan is required to produce better data in the future.
The New Zealand Hydrological Society is showing leadership in developing a path forward. A ratings workshop in Christchurch on the 11th and 12th of April was motivated, in part, by the publication of the National Environmental Monitoring Standard (NEMS) for the construction of stage-discharge and velocity-index ratings and, in part, by discussions with respect to the International Streamflow Rating Curve Project (ISRCP).
The workshop started with an international perspective of how ratings are managed in the United States, Canada, France, Australia, and New Zealand. The rest of the workshop explored many of the issues that need to be well understood to guide development of a best practice approach. All workshop presentations are available online on the NZHS web site.
This may be the first, but it definitely should not be the last international workshop on ratings.
There is interest in holding a similar workshop in Australia and I will be promoting the idea of North American Stream Hydrographers (NASH) hosting a ratings workshop at the CWRA conference in Lethbridge Alberta in 2017. It would be best if we can also inspire similar workshops in Europe, Asia, South America, and Africa. Please let me know if you would like to be involved with organizing a workshop in your region.
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. Read whitepaper here.