What pathways does sediment take as it travels through an estuary? Yesterday, Laurie van Gijzen defended her thesis, entitled “Sediment Pathways and Connectivity in San Francisco South Bay“. Laurie is one of the master’s students that I supervise, and she has done a great job on this project.
San Francisco Bay is a massive estuary, with over six million people living nearby. In addition to San Francisco, Silicon Valley sits on its shores. Some of the biggest tech companies in the world like Google and Facebook have their head offices right next to the Bay. For over 150 years, the ecological health of the bay has deteriorated, in part due to land reclamations and contaminated sediment from gold mining. The dynamics of San Francisco thus have a huge economic, social, and environmental impact.
Laurie’s work focused on calibrating and improving a sediment transport model of the bay, in order to track the pathways of fine sediment (i.e., mud). She worked with a notoriously fickle model (DELWAQ) and succeeded in greatly improving its calibration.
Laurie’s thesis summarized into a single diagram (Figure 6.1 from her report). She shows the dominant sediment pathways as dark arrows, and the net accumulation (import, in orange) or depletion (export, in blue). Also indicated are the dominant physical processes responsible for sediment transport in the different parts of the bay. The baroclinic processes mentioned here are currents resulting from density differences in seawater due to changes in salinity and temperature.
Another cool thing about her work is that Laurie was the first person to apply the coastal sediment connectivity framework that I have been developing! She was able to use this to identify key transport pathways and critical locations in the bay. It was extremely helpful for my research, as it gives us a proof of concept that our framework is applicable to multiple sites and can tell us something useful.
Her work was also accepted for a presentation at the NCK Days conference, which was meant to be held this week in Den Helder, but was cancelled due to ongoing societal chaos. Great job, Laurie!
It was Christmas 2016, and I felt like I had bitten off more than I could chew. I’m not talking about turkey, though. Four months into my PhD, I was feeling completely overwhelmed and starting to wonder what I had gotten myself into.
The goal of my project is to identify the pathways that sand takes as it moves in and around the Wadden Islands in the northern part of the Netherlands. Since the Dutch coast has a chronic erosion problem, accurately accounting for the whereabouts of their sand is a matter of national security. Right now, the Dutch deal with a deficit in their coastal sediment budget by adding more sand or “nourishing” wherever there is a shortfall.
Knowing when, where, and how much sand to add is especially challenging around these islands. Here, the persistent push and pull of the tide competes with the chaotic brutality of the waves to move sediment in complex patterns. These patterns are hard to predict with our usual box of tools, so we planned to throw everything we had at the problem: state-of-the-art field measurements, sophisticated computer models, reams of historical data, and a support team of experts from across the Netherlands. As PhD students go, I felt [and still feel!] pretty darn lucky to be a part of such a large and well-conceived project.
The Spaghetti Problem
However, as I started reading more and more about my topic, my initial enthusiasm began to wane. I was floored by just how much research had already been done on what I had thought was a fairly specific niche. The Dutch have been scrutinizing their coast for centuries, and to my inexperienced eyes, it seemed like they had already thought of everything.
There was another problem: at the end of almosteverystudyaboutsedimentpathways, there seems to be a diagram summarizing all the paths with lots of curvy arrows flying all over the place. This veritable plate of spaghetti makes for a nice conceptual drawing, but how can you statistically compare two plates of spaghetti with one another? A “past spaghetti” and a “future spaghetti”, to help understand potential responses to climate change? A “Dutch spaghetti” and an “American spaghetti”, to make my findings more general and useful for other places? If I was going to get anywhere with my PhD, I needed a spaghetti system.
Mmm, sediment pathways… Also, to be clear, this is not The Magical Figure That Single-handedly Changed My PhD. [Source: Flickr]By Christmas, I felt like I was in a weird purgatory between “it’s all been done before! I’ll never come up with anything original!” and “this is insurmountably complex and you’re foolish to think you’ll ever figure this out”. And just a dash of “how-did-I-get-here?” imposter syndrome, for good measure. I spent much of my holiday feeling overwhelmed and inadequate, like I couldn’t possibly live up to my own expectations, or (what I thought to be) the expectations of those around me.
But: new year, new start. On January 11th, 2017, my first day back in the Netherlands from holidays, a paper about coral reef hydrodynamics popped up in my Google Scholar alerts. At that time, I was also finishing up a paper about predicting floods on tropical islands, and I liked to keep an eye on the latest developments in that topic.
“A coupled wave-hydrodynamic model of an atoll with high friction: Mechanisms for flow, connectivity, and ecological implications“. Sounds promising, I like wave models.
In this paper, they wanted to understand how waves and ocean currents move water around Palmyra Atoll, a coral island in the middle of the Pacific. Coral reefs all around the world are in big trouble, and to help them we must first understand the physical processes governing the life and death of corals.
Palmyra Atoll, a coral island in the middle of the Pacific. Unfortunately not my PhD study site… [Source: Wikipedia]This was all very interesting stuff, though not particularly relevant to my research about flood prediction, since they seemed more focused on the ecological impact of their results. It was seemingly even less relevant to my PhD topic on Dutch sand- stay focused and stop wasting your time, Stuart! But then I turned the page and there it was:
The Magical Figure That Singlehandedly Changed My PhD…
The Magical Figure That Singlehandedly Changed My PhD, Figure 13 from Rogers et al (2017). “Connectivity between hydrodynamic zones. (a) connectivity matrix showing the probability a water parcel passing through a destination zone came from a given source zone, and (b) geographic connectivity of top 10% of pathways, where shading is relative importance as an overall source, width of line is relative strength of connection.”
Essentially, the authors had summarized the pathways that coral larvae can take around an island in a mathematically elegant way. This was pretty much identical to the goal of my PhD, if you substitute coral larvae for sand, and an idyllic Pacific island for a stormy estuary in Holland. They did it with a concept called “connectivity”, and it became immediately apparent that I had some homework to do.
So what the heck is connectivity?
So what exactly do they mean by connectivity, and how are we meant to interpret that magical diagram? Let’s start at the top. The upper panel is what we call an “adjacency matrix”, but you can think of it just like one of those mileage charts that you sometimes see in the corner of highway maps.
A mileage chart, which you can read in the same way as the adjacency matrix above. If you want to get between two points, just find the intersection between your row and column of choice. 290 miles from Birmingham to Edinburgh doesn’t sound too bad, until you encounter the menace that is Birmingham traffic… [Source: The Open University]Instead of looking at the distance between two points like in a mileage chart, the authors of the reef paper consider the likelihood of water travelling from one point to another. Darker squares show a higher chance of connection, and lighter squares, a lower chance. For instance, if we look at the first column, water is more likely to flow from the point they call “WT FR NW” to point “WT W” than it is to flow to “WT FR SW”.
The second panel shows the same information as the matrix, but this time actually showing the connections on a map – a “network diagram”. The thickness of the blue lines on the network diagram indicate how strong a connection between two points is. If all this seems rather familiar, then that’s probably because you’ve already met our network diagram’s more famous cousin, the transit map:
Harry Beck’s famous map of the London Underground. This map is a network diagram, much like the coral atoll connectivity map shown earlier. It represents the stations as nodes, and the train lines between them as links. It shows the connections symbolically, rather than at their true geographical locations, but this makes it easier for use to focus on the important things, like knowing where we need to switch train lines to get back to our hotel. [Source: Transport for London]
Cool maps. So what?
After seeing the coral reef connectivity diagram, I started googling and soon realized that I had been woefully ignorant of an entire mathematical discipline. Network theory represents complex systems as a series of points and the links between them. Once you’ve done that, interesting patterns start to emerge, such as the “six degrees of separation” or “small-world” phenomenon. It has been used in neurology, sociology, ecology, epidemiology, geomorphology, and basically every kind of “-ology”, except for coastal science and engineering. As far as I can tell, we seem to be the last ones to the party.
The more I read, the more excited I became, and the more vital it seemed for our field to catch up. Connectivity could help us quantify and bring order to the chaotic spaghetti churned out by our models and measurements – if we could figure out how to adapt it.
The course of my PhD was changed instantly with the discovery of that figure. Not only did connectivity provide a potentially useful tool, but it jolted me out of my funk and got me excited about my PhD again. It was an important finding for my research but not a “eureka moment” where everything was suddenly solved- far from it. It has been a long uphill slog since then, but with the help of some very clever people, I think we have almost reached our first milestone. We presented our early findings at a conference in 2017, and right now we’re in the final stages of preparing a scientific article about our ideas. That paper will then have to survive the woodchipper of peer-review, so it may still be many months before my work sees the light of day. But I remain hopeful.
Would I have stumbled upon connectivity eventually, had I not seen The Magic Figure? Probably not if I had only stuck to reading papers about coastal sediment transport. This finding has shaped my attitude towards coastal engineering research- I believe that the next advances in our field will not come from developing a new bedload equation, but from adopting new tools and techniquesfrom other disciplines. Not that we don’t need better bedload equations – I just don’t think I’m the guy to do it, and I think that we could all benefit from looking over the fence at our neighbours in other fields from time to time. As William Zinsser nicely put it:
“Think flexibly about the field you’re writing about. Its frontiers may no longer be where they were the last time you looked.”
The main protective barrier for the Netherlands against the threat of flooding from the sea is a row of colossal sand dunes and wide beaches that stretch the length of their coast. However, that barrier is not completely natural — since the Dutch coast is in a constant state of erosion, the sand in their coastal zone has to be continually replenished. This replenishment takes the form of nourishments, which are essentially just massive piles of sand placed on beaches, dunes, or just offshore. The Dutch are lucky, since the bottom of the North Sea is covered in sand for hundreds of kilometers in every direction, meaning that there is a ready supply available for this purpose.
Although we still have plenty to learn about how to construct these nourishments effectively and in an environmentally friendly way, we are starting to get the hang of it — at least for long, straight, sandy coastlines like in Holland. However, this all gets a bit trickier when we turn our attention to the Wadden Islands dotting the northern coast of the Netherlands. These little islands sit between the stormy North Sea and the shallow Wadden Sea, a large estuary whose ecological value is unmatched in the Netherlands.
The coast of these islands is punctuated by a series of inlets connecting the two seas. Chaos reigns at these inlets, where strong tidal currents pass in and out, clashing with waves and whisking sandy shoals in and out of existence in unpredictable ways. This makes the inlets treacherous for ships, but also a challenge to simulate with our computer models and design nourishments for.
How, then, are are we meant to nourish the coast of these islands? We want to keep their inhabitants (and those on the nearby mainland) safe from flooding, but also need to be careful about inadvertently disrupting the vital ecological habitat of the Wadden Sea.
To answer that question, the Dutch government initiated the Kustgenese or Coastal Genesis project. In collaboration with several Dutch universities, companies, and research institutes, they set out to better understand how these tidal inlets work, and whether it is possible to effectively nourish them. The project focuses on Ameland Inlet, which is located between the islands of Ameland and Terschelling.
My PhD project is but a very tiny piece of the very large Kustgenese pie. My goal is to figure out specifically how the size of sand grains affects the paths that they take around tidal inlets. It has been the dream job for someone who has loved playing in the sand ever since he was a little kid. As a result, it has entailed a lot of time at my computer and in the laboratory, investigating the characteristics of the sand in Ameland inlet (that’s also why I have so many pictures of sand on this blog- we have a really cool microscope!). It is very fine sand and would be absolutely perfect for squidging your toes through on a hot day — if it weren’t at the bottom of the sea, that is:
Native sediment from Ameland ebb-tidal delta
In the spirit of ‘why not?’, the Dutch government decided that the best way to test whether nourishments would be effective in this environment was to just go ahead and try one out last year. They dredged up 5 million cubic metres of sand (that’s enough to fill 3 Skydomes, for anyone reading this back home in Toronto) and placed them just outside the inlet. A few months ago, one of the Dutch government officials showed up at a meeting with a “present” for me… some sand from the nourishment!
Nourishment sediment dredged from offshore and placed on Ameland ebb-tidal delta.
Needless to say, I was very excited. At first glance, it appears quite similar to the native sediment, so that means it should behave in a similar manner. Time will tell how the nourishment evolves- we are watching very closely!
I have spent many months holed up in the laboratory counting green grains of sand. Last year we dumped over 1 ton of fluorescent, magnetic tracer sand into the North Sea, where the waves and tides then scattered them along the coast. We then spent the following weeks circling around on a boat to try and find it all again. We scooped up over 200 samples of sand from the seabed, then brought them back to the lab for analysis. We used a super strong magnet and blue UV light to separate the tracer (bright green) from the normal sand (looks grey or purple under UV light). This part is REALLY boring because most samples don’t have any tracer but we still have to look hard for it . But then we get to look at all the sand under a fancy microscope, which is my new favourite toy! Under the UV light, the tracer reminds me of little green constellations of stars in a purple night sky.
The tracers glow bright green under ultraviolet lights. On the top left we have a jar of normal beach sand sitting next to a jar of tracer. The distinction becomes clear once we place them under UV illumination (top right). This is especially important when we analyze samples taken from the seabed, where there may be only a few grains of tracer (bottom left). The fluorescent properties of the tracer help it stand out from normal sand (bottom right), which lets us count the individual grains.
Fortunately, we can use computers to count the individual grains and tell us their size. With this information, we can estimate how the size of a sand grain determines how far and fast it will travel. This is important for planning sand nourishments to protect the coast.
If we can figure out where the green tracer did (or didn’t) go, that will tell us how normal sand moves around on the Dutch coast. And this will hopefully keep our feet dry here in Delft for a long time to come!
Coastally Curious 