Timely, given the beaver controversy, that I had actually spent a few hours on Friday drawing aerial diagrams of my recent river project to animate a presentation I’m planning on the possibilities of ambitious chalk river restoration.
This project is the latest phase of a piece-by-piece catchment restoration of the River Nar, funded this time by a Water Environment Grant and managed by myself as designer and the Norfolk Rivers Drainage Board as project facilitators (they did all the difficult stuff like Health and Safety, managing the finances etc.) particularly Paul George, Helen Mandley and Caroline Laburne.
The primary driver was the creation of hydrological or lateral connectivity around a part of the river that had historically been diverted to the edge of the floodplain to build up a head of water to drive a mill. This type of channel diversion is very common on chalk streams, but it has the effect of divorcing the river from the floodplain, and lowering the gradient over long reaches in order to build up the head at the mill. More or less universally this creates a sediment retention issue, especially nowadays that the mills are not worked.
Add late 20th century dredging or zealous weed management to the mix and you end up with a channel that flows slowly, is full of sediment and burr-reed, cannot scour itself clean and is disconnected from the floodplain.
The drawings below show Phase 1, the uppermost part, which we carried out in late 2019. We finished Phase 2 in autumn 2020. We came in under budget, so hopefully now there will be a Phase 3.
The top diagram shows the river before the works, the red lines marking the edge of the floodplain. You can see how the river has been diverted to the very edge of the northern (left) part of the floodpain. With the river thus diverted and also overdeep as a result of dredging, the option to restore the gravel bed of the river in situ would have been enormously expensive, unsustainable and not ideal anyway, given that the channel was kind of in the wrong place.
I designed a new channel planform flowing roughly where the river would once have been (the very upper hundred yards actually cuts a corner, but it does that because only at that point could I come off an undamaged part of the original channel bed and start with a good baseline level). I based the meander shapes on field measurements of as many local original meanders as I could, including some lost original meanders.
I calculated the gradient from the very uppermost point to the return point which would be 2000 meters downstream. This phase illustrated is about 450 metres. Based on the number of meanders at the estimated wavelength I ended up with a fairly easy to measure 5cm fall from one inflection point to the next. Between these we cut the bends about 30 cm deeper on the outside below the undercut bank. We also ended up with piles of gravel here and there from the channel excavation and these we tossed back in at the end to give the river some grit to work with.
Look carefully at the diagram and you’ll see we only cut off the top fifty yards of the old channel. It still has flow: it is spring-fed from groundwater seepage and some quite significant springs at the corner. So, to add to the channel types and biodiversity of the floodplain as a whole, we put a high bar of gravel just upstream of where the new channel runs back in: this has backed up the water in the old channel and created a more fen-like habitat, or pseudo oxbow.
Finally, a few months after it had all settled in we pinned a very large number of trees and branches to the new river bed: not as many as I would have liked to, now I go back and take a look. I plan to do more in the new channel and the old, to get as much complexity in there as possible and in the old channel to pack it out and saturate the floodplain.
Early results from invert and fish surveys courtesy of the Norfolk Rivers Trust look really good: a healthy size range of trout (lots of juveniles and a few old snorters) and good invert numbers and diversity.
I’ll write about the next phase in the next blog post, when the overall design will be made clearer, especially why we had to run this to phase back into the old channel. In the next phase we end up with three channels side by side and some pseudo beaver ponds too. And a short “Stage Zero”.
The pictures below the diagrams probably say more than lengthy explanations can.
The sequence above shows the same bend as it is cut, then in the autumn 2019 just after the works had finished (top right) spring 2020 (next one down) summer 2020 (bottom left) and one year on, autumn 2020 (bottom right).
The sequence above also shows several shots of the same reach, just as we are tidying up (autumn 2019) (note the pile of LWD ready to go in), summer 2020 (bottom left) (we have Konik ponies to help manage the floodplain), autumn 2020 (middle right) and under high flows winter 2020.