Bed load: as inscrutable as ever

 

One of the many arguments in favour of good old-fashioned libraries – the bricks and mortar, hardcopy, poking-around-in-the-stacks  variety – is the un-measurable element of serendipity: things you happen upon when looking for something else. I was recently relishing the joys of the Geological Society Library here in London, in quest of material for a post that I have yet to write, when exactly such a diversion occurred. Taking down from the shelf the recent issue of Sedimentology that I was after, I naturally scanned the contents and was struck by a completely different title: “The partitioning of the total sediment load of a river into suspended load and bedload: A review of empirical data.” It admittedly does not sound particularly gripping, but I knew it was a tricky geological conundrum and thought that perhaps, finally, here would be a definitive breakthrough; my hopes were quickly dashed, but the paper did fulfil my enjoyment of the apparently simple things that remain elusive, that continue to defy our best attempts to measure and quantify them.

The news today is filled with dramatic and tragic images of rivers in flood, devastating lives and livelihoods. Rivers are engines, whose dimensions and power increase by orders of magnitude when they hurtle out of control. And, in doing so, they are gargantuan conveyor belts of sediment transport, shifting unimaginable volumes of material from the continents to the oceans – along the way, ironically, renewing the fertility of agricultural land after first devastating it. The engine operates in three different ways to carry its sediment load: by rolling and bouncing grains, pebbles, and boulders along its bed, by swirling smaller grains into almost continual suspension, and by carrying soluble minerals dissolved in the water. The last, the dissolved load, is simple to measure, the suspended load relatively simple, and the bed load extremely difficult. Measures of the sediment load of rivers commonly are for the suspended load only, but the bed load is huge. And all of these numbers are only averages—a river in flood will carry orders of magnitude more sediment than it does on a normal day.

Understanding the sediment load of a river and how it works is much more than simply a matter of scientific enquiry – it’s vital for managing soil erosion, infrastructure design, reservoir sedimentation, and agriculture. Yet we are not very good at it. As I mentioned, the dissolved and suspended loads are relatively easy to measure, the bed load extremely difficult – and, in times of flood, downright hazardous. Yet the bed load is critical. A few years ago, I stood next to a river flooding out of the Pyrenees (the photos at the head of this post show its normal, tranquil, state and in flood). Behind the roaring cacophony of the water racing by was a deep undertone, a thundering and clunking sound: this was the voice of the bed load, of boulders and cobbles rolling and bashing their way along the river bed. Floods in this part of France are common and destructive – the photo below shows a river’s bed load deposited in the town of Vernet les Bains (ironically, “Vernet the Baths”) after what was possibly the European rainfall record, a metre in 24 hours in 1940.

 

Historically, the challenge of measuring bed load and its dramatic variability has been approached empirically, by attempting to devise a relationship between the (measurable) suspended load and the associated transport along the bed. Albert Einstein’s son, Hans, produced the classic equation, still used with modifications today, when he was a professor at the University of California, Berkeley (his father had advised him against embarking on this field of study because of its complexity). In the early 1950s, several pieces of work produced empirical tables listing bed load as a percentage of total and suspended load under different conditions, particularly the material of the river bed; these tables, reproduced in the paper I discovered in the library (full reference below), look like this:

 

As an attempt to cram a wildly variable natural phenomenon into some kind of quantifiable and useable scheme, they are obviously of only limited use – just look at the variability in the percentages even when a number of variables are ignored. I read the start of the abstract of the paper with interest:

The partitioning of the total sediment load of a river into suspended load and bedload is an important problem in fluvial geomorphology, sedimentation engineering and sedimentology. Bedload transport rates are notoriously hard to measure and at many sites only suspended load data are available. Often the bedload fraction is estimated with rule-of-thumb methods such as Maddock’s Table, which are inadequately field-tested.

But then my expectations were dampened when I read the abstract’s conclusion:

 For sand bed rivers, the bedload fraction may be substantial (30-50%) even for large catchments. However, available data are scarce and of varying quality. Long-term partitioning varies widely among catchments and the available data are currently not sufficient to effectively discriminate control parameters.

In other words, we’re effectively no closer today to tying down and understanding a river’s bed load than we were more than fifty years ago. However, the paper is a really interesting review of the problem, and it adds some new high-quality data from a river in the mountains of Austria, “for which high temporal resolution data on both bedload and suspended load are available.” But, “The available data show large scatter on all scales.”

And here, just to illustrate the mind-boggling range of measurements of how rivers behave, are two graphics from the paper. First, bed load transport rates as a function of suspended load transport rates (in kilograms per second – note that the scales are logarithmic):

 

and then a plot of the fraction of the total load that is transported as bed load (this varies between zero and one) versus the associated suspended load concentration (which varies between 0.3 and 10,000 parts per million). Random scattering, anyone?

 

OK, it’s not random, but for me it’s a fascinating example of the elusive and complex character of apparently simple natural processes – and ones that are important to us.

[The reference for the paper is: Turowski, Jens M.; Rickenmann, Dieter; Dadson, Simon J.. 2010 The partitioning of the total sediment load of a river into suspended load and bedload: A review of empirical data. Sedimentology, 57. 1126-1146. doi: 10.1111/j.1365-3091.2009.01140.x]

Comments

[F]

I find this information (and lack thereof!) about riverine systems increasingly interesting as I continue to read of such things. Thanks to your blog, Anne at Highly Allochthonous, and Brian of Clastic Detritus, I have a wealth of information and pointers to further information that would be very hard to come by if paying for peer-reviewed literature was the only access available. All praise to the internet, and the people who provide worthwhile content.

[Brian Romans]

Michael, Thanks much for posting this. I'm actually grappling w/ this problem for a study I'm doing right now on the Santa Clara River (and nearby rivers) in southern California. We've done some paleo-sediment budget calculations using cosmogenic nuclide-derived bulk erosion rates from watersheds compared to depositional rates in the offshore sinks for the Holocene. The whole suspended only-load vs. entire load for historical measurements is rearing its ugly head. An interesting problem indeed.