A lot of my research work at University of Southampton is related to the effects logjams in rivers have on flood behaviour. The recent heavy rain and flooding in the South provided me with an opportunity to get out into the New Forest and have a look at some logjams during extreme river flows.
The main process by which logjams (or indeed any fallen tree) in a river affect the hydrology and hydraulics is through increasing resistance to flowing water. In simple terms as water flows past the obstruction it “uses up” energy due to increased turbulence, this is most pronounced if there is a “step” in the water profile; this means the river is behaving as if there is a weir or a waterfall, or perhaps a length of rapids.
Although water seems to be moving faster when it is spilling over a weir the net result is a slowing down; water “ponds” upstream of the obstruction forming a slow moving backwater and sometimes even a mini-lake, it then moves very quickly over/under/through the obstruction before slowing down again downstream. If we imagine point A some distance upstream and point B some distance downstream with a logjam in the middle the water will get from A to B slightly slower with the logjam there than if the channel was clear, even though in places it will be moving really quickly.
So how does this affect flooding? Well there are two main results, the first is that upstream of the logjam, and potentially for a short distance downstream the river is likely to over top its banks and flow onto the floodplain, perhaps extensively. The photo to the right shows how water can pond upstream, over top the banks and then flow in shallow, temporary floodplain channels around the jam (central photo above) and then re-enter the channel. It is important to emphasize this is a perfectly natural process with many benefits and will cause no problems provided the land that is inundated does not contain important infrastructure (houses, roads, etc). These wet floodplains can be important temporary habitats for a variety of species as well as a “sink” for fine sediment in the river. River water typically contains fine sediment carried along in the flow, as it moves over a floodplain during a flood the water slows down and no longer has enough energy to transport the sediment so it drops out of flow onto the floodplain. This process of floods depositing nutrient-rich sediments has been the cornerstone of lots of fertile agricultural land (e.g. Nile, Mississippi) since at least the stone age.
Now I’ve mentioned that a logjam can slow down the passage of water across it, but what effect might this have elsewhere? Well this is a very complex picture. By and large slowing water down during a flood event and enabling it to flood uninhabited floodplains is a good thing, however the combined effects of lots of these individual mini effects (lots of logjams all over the place) is not clear cut.
If we imagine cars as little packets of water and rush hour as a flood, we can see if we delay or slow down a few cars then the congestion at rush hour generally will be a little bit better. And we’d think that if we slow down more and more cars eventually all traffic would flow smoothly (i.e. no flood). However if by mistake we slow down some cars that are already home by 5pm we might delay them so they are additionally part of the rush hour making things even worse; for this we could imagine slowing flood water down near a town so that it reaches the town at the same time as the bulk of the flood wave, rather than passing through before the flood wave arrives.
For this reason careful study at a catchment scale is needed to understand how floodplain land use feeds into flood behaviour to pick up interactions and feedback loops of the sort I mentioned above in my laborious rush hour analogy! This is the motivation behind research work I’m conducting at University of Southampton in association with the Environment Agency. By working out where water can be slowed down and given space to flood we may be able to reduce flood risk downstream at vulnerable urban locations. Such an approach can work alongside traditional engineered flood defenses, and potentially in locations where flood defenses are not economically viable (such as protecting small villages).