Room for the River – Dutch flood control

Maeslantkering barrier – Netherlands
© World66 / Wikimedia Commons / cc-by-sa-1.0

As I detailed in my previous post about the current flooding in the UK, the main focus at the present time should be on protecting lives and property. It is certainly not the time for blame and is too early to attempt to learn lessons from the flooding or the response to it. That said I’ve seen quite a few articles and comments that the UK needs to learn from the Dutch in terms of flood defence. This seems an especially compelling argument in the case of the Somerset Levels where there seem to be many parallels in a tract of land close to, or below sea-level and criss-crossed by artificial drainage channels and pumping stations. With that in mind local MP Ian Liddell-Grainger visited Strasbourg two weeks ago to talk with Dutch politicians. I happen to agree with this assessment, we can learn a lot from the Dutch; however I think there are a lot of misconceptions about exactly what Dutch flood risk management entails. I thought it would be interesting to take a more detailed look at flood risk management in the Netherlands.

It seems as if the perception of the Dutch water management programme within the UK is stuck in the pre-1990’s. People seem to believe that hard engineering is still king in Holland and that the river and sea is forced back and kept out. Nothing could be further from the truth. I hope to give a brief look at how the Netherlands is responding to climate change in respect of flood risk and show that this is in no way as simple as many UK commentators would have you believe.

I think most people would probably be aware that the Netherlands consists of large tracts of reclaimed land, which is often below sea level, and by as much as 6-metres in place. It is effectively part of the delta of the Rhine-Meuse river system. Much of the land is protected by an extensive system of dikes, creating tracts of land called polders, which are farmland below the level of sea/rivers and which rely on pumping stations to keep them drained and to move flood waters off the land (in a similar way to the Levels). This extensive network of man-made drainage and reclaimed land which was started in Roman times leads to the nice sayingGod created the world but the Dutch created Holland”.

The recent storm surge on the East Coast of the UK, accompanied by dramatic pictures of the tidal barrier on the River Hull nearly overtopping were compared to the most devastating storm surge of recent times, that of 31st January 1953. In Holland this same storm surge is called simply “Watersnoodramp” (The flood disaster); whilst 326 died in the UK, 2551 people lost their lives in Holland. This disaster led to an overhauling of the Dutch approach to flood defence including the construction of the iconic Maeslant barrier (pictured above).

It is this image of the Netherlands as a nation with impregnable hard engineered flood defences and water management that make sure water goes where it is supposed to and does what it’s told that people have in mind when they advocate learning lessons from Holland or getting Dutch flood defence expertise into the UK. However this image is out-dated. For over ten years the Dutch have been embracing a new approach to flood risk management marrying the best of their engineering expertise with more flexible management of flood waters.

The reason behind this rethink is climate change; over time the Dutch have been building their dikes higher and higher but realised after flooding in the 1990’s that higher dikes meant more devastating floods when they failed. Not only that but drained land behind the dikes is subsiding (and is thus even lower). There is an explicit recognition in their policy that rainfall is and will be harder and more frequent. They have therefore begun to instigate managed retreat of some land, along with upgrading, rebuilding and improving flood defences. One example of such a project is Ruimte voor de rivier (Room for the river), the objective being to give the river at 30 locations more room to accommodate higher water levels, essentially allowing the river to flood safely.

Work includes the dredging of river channels but this is only a small aspect of the plan. In Nijmegen a new relief channel is being cut for the River Waal 200m wide and 3km long, along with moving back the existing dikes; this has involved the demolition of 50 houses. The floodplain will be park and recreational ground to allow safe flooding. This is a huge change to a 2000 year old city.

Plan for Overdiepse Polder designed by Bosch Slabbers – farms are demolished and rebuilt on large mounds.
© Bosch Slabbers (

Another component is “de-poldering”; lowering/removing the dikes in some areas of farmland and designated in the land as a flood zone. The inundation of these former agricultural areas during flood events will protect towns elsewhere. In other areas retreat is being managed by protecting farm houses and acknowledging the farm land itself will flood. This is being done by demolishing the existing farm houses, constructing giant man-made mounds (~20 acres each) and rebuilding farm houses on top of them (hopefully) higher than the flood waters will reach (~7-8m high). One example is Overdiepse Polder, where 16 farms have been designated as flood zones to protect 140,000 residents of Den Bosch. As the New York Times put itBy displacing farmers.. residents in that city can breathe a little easier.” Individual houses are also increasing being built to be flood resilient in novel ways, including building floating, or amphibious houses.

A new floating home installed by Deltasync in Delft, Netherlands, Dec ’13.
© Deltasync (

So, can we learn from the Dutch about flood control? Undoubtedly. In certain respects they could be said to be 10-15 years ahead of the UK in aspect of flood risk management. In the case of the Somerset Levels, experience from Dutch water managers could be very helpful. However, it is disingenuous to cherry pick one small part of a comprehensive and resilient flood risk management plan such as “room for the river” (dredging) and suggest applying it would solve UK flood problems. The “Dutch approach” to flooding specifically on the Somerset Levels would likely involve widening and deepening the Tone & Parrett (through dredging) as well as abandoning large tracts of farmland as flood zones and the demolition and relocation of existing farm houses onto large mounds. Make no mistake such a policy was not immediately popular in the Netherlands, as I suspect it would be unpopular in the UK.

Another consideration is money, the Dutch have earmarked €100bn through to 2100 to cope with sea-level rises of 1.3m. This is for 451km of coastline. The UK has a longer coastline than countries such as Brazil and India at 12,429km. If we were to match the Netherlands € for €, km for km, that would be about €2.8 trillion by the end of the century purely on coastal defence (no river defence). This is obviously a silly example, as much of our coast is hard cliffs which need no defending, but it does illustrate the vast gulf in scales between Holland and the UK. It also illustrates the current spend of €0.67bn a year for all flood defence (coastal & inland) is woefully inadequate.

The most important advice we can take from the Dutch experience of flood risk management is that you need to be adaptable, open minded to changing historic land use patterns and prepared to take radical solutions. The climate is changing and flood risk management practises will have to change with it.

Posted in Flooding, Geomorphology, Hydrology, Politics, River Management | 9 Comments

Scramble for blame serves no-one


Flooding on Somerset Levels (copyright Kevin Keatley

The flooding blighting much of the UK has dominated the news now for two weeks and shows little sign of abating for the next 7-10 days. Initially when the flooding was fairly isolated to the Somerset levels the reporting in newspapers and on television was led, and dominated by, science and environmental commentators who took pains to research their articles and base their conclusions on the best available evidence and advice.

As the flooding story continued, and became more of a national political debate, the coverage has increasing become dumbed down to human interest stories and the excellent reporting from aforementioned science correspondents has been buried beneath coverage of train cancellations and people evacuating their homes. This has had the pervasive effect of needing to condense the scientific explanations of flooding into easily digested throw-away comments in reports. The hugely successful FLAG local campaign group in the Levels have been campaigning for a long time for increased dredging of the Rivers Tone & Parrett. The media love a human interest story and the combination of distraught people, large gatherings with placards and a simple explanation for flooding (which also happens to have a simple scapegoat) proved irresistible.

The cumulative outcome of this is that the human interest stories on the Somerset Levels dominates the news, and the narrative that has emerged is one of a lack of dredging leading to flooding. The government, against advice of the Chartered Institute for Water & Environmental Management, leading hydrologists and flood scientists, has embraced the dredging narrative and committed to dredging the levels. There are two fundamental problems with the way this issue has played out in the media.


Mulcheney, Somerset Levels (copyright flickr user: nicksarebi

Firstly it is highly doubtful dredging rivers in general, or the Tone & Parrett specifically, would have any appreciable effect on flood events of the magnitude we have witnessed recently. To use one of my trademark analogies, it’s a little like drinking two bottles of wine and suggesting the brandy chaser was the thing that made you drunk. It’s possible it might be the case, but it’s highly unlikely. Stretching the analogy (as is my wont) and speaking purely of my own alcohol tolerance, I could drink 1 or 2 glasses of wine, with or without a brandy and be coherent. If I drink 4 glasses of wine, I’ll likely be hammered either way. It is only in the very specific case of my consuming precisely 3 glasses of wine that drinking, or not drinking a brandy will have any difference whatsoever. That is the essential problem with dredging; increasing the river conveyance only has an effect for a very specific flood event which would just over-top the un-dredged river (and thus may not over-top in the presence of dredging). It’s a little more complex in managed systems like the Levels, but the magnitude of event is most relevant issue. This one was far, far too big.


FLAG confronts Owen Patterson MP (source: Facebook)

Secondly, and most importantly, this issue is completely irrelevant at the present time. The priorities in a natural disaster should be firstly to prevent or minimise loss of life, prevent or minimise damage to infrastructure and to ensure as quick a recovery as possible after the event. In this respect the relevant bodies are doing their best to deal with the life and property issue. We should already be talking about practical measures to help people clean up after the floods and to provide counselling and support to those affected. This is how the World Bank is helping Malawi deal with climate change and flooding. This should be the dominant narrative. Instead by focusing on who is to blame for a natural disaster we risk being ill prepared for the challenges when the waters recede. At this stage it is completely irrelevant who is to blame for flooding (and for the record I don’t believe anyone, or anything IS to blame). Apportioning blame serves no-one at the present time. In this respect the FLAG campaign on the Levels has been somewhat a victim of its own success. They have successfully steered the national debate on flooding towards drainage and dredging in an area where only 40 homes have been flooded (compare to 48000 homes and £3-6bn damage in 2007 floods). I say these figures only for comparison, not to belittle the huge suffering of those affected in both events.

The problem is in 2007 the debate was on assistance, insurance payouts, flood funding and proper analysis of where future flood policy should go (culminating in the extensive Pitt Report). In short the problem was seen as a natural disaster and focus was on sorting out the aftermath and learning lessons (albeit with mixed success). In these floods by focusing on blame, the narrative seems simpler and has culminated in “dredge the rivers, stop the floods”. The government has seized on the scape goat with wild abandon, Eric Pickles recently savaging the EA on the Andrew Marr show and committing to dredging the Tone & Parrett and generally appearing to embrace a retrograde approach to risk management. So problem solved?

Well, no. The rivers will be dredged, the Somerset Levels will still flood in a similar sized rainfall event, to broadly the same depth, extent and duration, and none of this helps people get their homes, livelihoods and lives back together in the coming months. We’ve been having the wrong debate.

As I write the number of severe flood warnings (risk to life) are rising on the Thames, Worcester is beginning to flood and no-one can seriously think this is due a deficit of dredging on the Thames and Severn. In focusing over the past ten days on searching for scapegoats and apportioning blame in increasingly preposterous ways the government has been desperately trying to deflect attention from cuts to the EA, as if they are Nero blaming Christians for the Great Fire of Rome. Moreover the complete failure to grasp, or take the time to brush up on hydrology means they’ve been parroting ridiculous claims that flooding can be stopped. It can’t. We can mitigate it, can make sure we are prepared for it to lessen effects and speed recovery, we can even manipulate catchment hydrology to “move” or concentrate flooding in one place (farms/countryside) to lessen flooding elsewhere (towns). We cannot however lessen the rain.

Meanwhile the maligned agency has been working tirelessly to mitigate the damage of the current floods. The government should focus on supporting the one body that can, and is, equipped to do something to lessen the impact of these floods. The government should also be thinking about how to support people in the eventual clean-up. They should not be wasting time worrying about who is to blame. If necessary that can come later with a proper enquiry and report.

Posted in Uncategorized | Tagged , | 42 Comments

How wood in rivers affects flood risk

Look upstream at a logjam

UPDATE: This work was accepted for publication in the journal “Earth Surfaces Process & Landforms” on 12/02/2016.

One of the key objectives of my PhD research at University of Southampton was to try and assess how changing volumes of wood in rivers affects the likelihood and magnitude of flooding. The amount of wood in rivers can increase for various reasons, such as artificial insertion in the river for river restoration, river managers choosing to cease clearing natural deadwood from channels, or the growth/restoration of streamside (riparian) forests. We know that if wood is put into a river then flood water moves slower through/around it, and thus for a short distance downstream flood water will have a longer “travel time”. What no-one has done before is look at the effects upon flooding at a distant downstream location (such as a town) of changing the speed water moves through a small sections of the river catchment upstream of it.

My research in this area was sponsored by the Environment Agency in the UK. Currently my thesis is awaiting examination, and for various reasons the publication of a key set of results in the academic literature is not likely to appear until the later part of this year at the earliest. Given the current interest the subject with articles in The Guardian and BBC in the past few days I thought it timely to write a short blog post giving a précis of the main findings of a computer modelling study I conducted on the effects of adding wood to rivers as a flood control method.

I used a computer model (OVERFLOW written by Dr Nick Odoni) also used for the experimental “soft engineering” flood defence project at Pickering, N.Yorkshire mentioned in Roger Harrabin’s article for the BBC. This model allowed me to run thousands of model variations to look at the effects of changing wood in rivers at a wide range of scales, but also different spatial arrangements within a catchment. What I am interested in is the effect upon flooding at a downstream urban location when wood is inserted into rivers through the catchment in various arrangements.

I am concious I need to restrain myself as I could (literally and demonstrably) write over 70 thousand words on this subject!

The catchment I modelled is within the New Forest National Park and has fairly low slopes, it is around 100km² and contains around 100km of stream length. The outflow from the catchment is the Lymington River at the town of Brockenhurst.

The key points that emerged from the modelling study are:

  • Adding artificial logjams to stretches of river as short as 500m can change the height of a flood peak in a downstream town, but these changes are very small and have highly variable magnitude and directionality.

To unpack this point a little more – putting logjams into a river in some locations can increase the depth of flooding downstream.

  • Broadly speaking steep headwater reaches with wood in them had little to no effect on downstream flooding – in very simple terms this can be put down to the water flowing in steep streams moving faster and having more energy, and thus being less susceptible to being significantly slowed by additional wood. I found streams with a slope of >0.005m/m or greater showed little change in flood peak upon adding wood.
  • In middle to lower parts of the catchment changes in flood height at the town are observed, but there is little predictability in response. I.E. with my data I am unable to predict whether putting wood into a river channel around 3-15km upstream of a town will decrease or increase flood risk.
  • Generally as the length of river with wood in it increases, the magnitude of change in the flood peak height also increases – i.e. wood in more river channels results in bigger changes to flooding. However the direction of this change remains highly variable.

Meander bend during flood, looking downstream. River channel flows from left to right, before curving back towards top left. Water is flowing rapidly over the floodplain on the left, bypassing the bend in the river.

The overall conclusion is that just inserting wood and/or logjams into rivers within a catchment as part of flood control for a downstream location is highly unpredictable. The reason for this is that locally logjams force water out onto the floodplain, if this floodplain is grass or scrub then the flood water is capable of still moving fairly rapidly across it, indeed in some cases it can flow directly down-valley and bypass bends in the river.

The real benefits in “rewilding” for flood control comes from restoring floodplain forests. Complex forested floodplains dramatically slows water moving over them as they have an irregular surface covered by tree roots, upright tree trunks and dead wood. My model runs for scenarios where wood is inserted into the river and a forest is allowed to grow on the floodplain show substantial and predictable responses in downstream flood height.

  • Restoring floodplain forests to short sections of river of 1-2km can reduce downstream flood height by 1% after 25 years growth.
  • There is predictable pattern of response with floodplain forest in lower parts of a catchment (e.g. near a town), increasing flood height. However when used in the middle and upper reaches of a catchment a reduction in flood height is modelled.
  • As the extent of restoration increases the magnitude of change in downstream flood peak increases and displays a similar spatial pattern (near the town increases flood height, further upstream reduces flood height)
A map showing an example of a sub-catchment area. This one is 14.6% of the total area and if forested is predicted to lead to a 5.3% reduction in flood peak height after 25 years growth

A map showing an example of a sub-catchment area. This one is 14.6% of the total area and if forested is predicted to lead to a 5.3% reduction in flood peak height at the downstream urban location after 25 years growth

The most promising scenarios, and the real take home message is the restoration of floodplain forests to entire “sub-catchments” of the main catchment (a tributary of the main river and all streams draining to it) always decreases flood peak height after 25 years growth, and can have dramatic effects. If this is done for areas of 20-35% of the catchment reduction in flood peak height  of 10-15% are modelled after 25 years of forest growth.


Flow out of a sub-catchment/tributary is slowed, this translates the hydrograph for that sub-catchment to the right (see upper right of figure) i.e. moves it later in time. The water flowing out of this subcatchment would normally arrive at the town downstream during the green part of the main hydrograph (the flood peak). By moving later in time it arrives after the peak – the new hydrograph is shown by the dotted line, which is lower, later and lasts for a slightly longer time. Figure from Dixon, 2014

As a simple analogy during a flood many “packets” of water are delivered to the main trunk river from all its tributaries. If the delivery of a single large “packet” of water can be significantly delayed it will then arrive at the main river after the peak of the flood, and thus the main flood peak height has less “packets”of water in it and is lower. Imagine it as a tapas meal where you’ve ordered too many dishes, your table is full already, but thankfully the waiter brings your albondigas after the gambas pil-pil are finished and thus your table doesn’t overflow with tapas.

The two most important implications for flood control using river restoration or re-wilding are:

  • Applying wood and/or logjams on their own to short stretches of river of 1-5km has a highly variable effect on flood peak height at a downstream urban location of ±4%. The (well known) local flood wave attenuation effects of logjams do not always translated to reduced downstream flood risk. Given the difficulty in predicting the response before installation this is a highly risky approach and should probably be avoided unless there is extensive site experience/local knowledge/investigation beforehand.
  • The most promising (and practical) scenarios are to restore small headwater sub-catchments representing 10-15% of the total catchment area, where reductions of 5-6% in flood peak height can be seen after 25 years, with this reduction increasing to 7-8% after 50 years growth. If the area is increased to a large sub-catchment representing 25-35% of the catchment area reductions of 10-15% in flood peak magnitude can be seen after 25 years (with again bigger reductions as the forest ages and matures); although restoration of such large areas may prove impractical.

It is important to note that this modelling only looks at the speed of water moving through the river network and off hillslopes, it does not take any account of the reductions predicted in the amount of water reaching rivers through trees increase infiltration rates of rain/runoff into soils, as mentioned by George Monbiot last week. So we could expect to see even greater reductions in flood peak discharge downstream than predicted just from flood wave travel time modelling.

In conclusion although governmental and public interest in the concept of “rewilding” rivers for flood control is promising, it is important to recognise that the local effects of wood in rivers slowing flow can have surprising and counter-intuitive effects when looked at in the context of a whole river catchment. We need to do a lot more work in this area and in the meantime the insertion of logjams and dead wood into rivers for flood control should be used with caution and extensive site analysis.

Posted in Flooding, Geomorphology, Hydrology, River Management, River restoration | Tagged , , | 64 Comments

Erosion Control and river restoration success

I’ve done a fair bit of research on the streams of the New Forest national park and have seen first-hand a number of restoration projects as well as the results of many more that have been undertaken in the past ten years. Many of these projects have been great successes, but there are a few instances of projects which have had unintended results.

One example which I’ve kept an eye for the past four years or so is a project which has experienced some geomorphological side-effects mainly outside of the restoration area. Whilst this is a site specific issue, and I’ll explain why I think the geomorphology has adjusted in unintended ways, I feel the issue says something wider about the importance of fluvial geomorphology in river restoration.


Erosion Control November 2011

The area in question is on a small river in the North-East of the New Forest national park and is on a meandering section immediately upstream of a small bridge where a forest access track crosses the river. The channel in this area is heavily incised below the floodplain by up to two metres and the banks are being actively eroded on the outside of the meander bends. In the picture shown the river managers have attempted to counter this “run away erosion” by installing staked logs as rip-rap to reinforce the banks. This photo was taken in November 2011. The picture below shows the same area earlier this month (November 2013).

Same bend in November 2013 where erosion control has failed

Same bend in November 2013 where erosion control has failed

The logs have been eroded away and transported downstream by some distance, considering the logs are 5.5m in length and around 30cm in diameter.

Looking upstream, log in foreground has moved around meander, approximately 30m from starting position

Looking upstream, log in foreground has moved around meander, approximately 30m from starting position

In order to consider why these bank protection measures have failed and what could possibly be done to arrest the erosion it is vital to consider the geomorphological context that this reach sits in, as well as the geomorphological history of the catchment. This is a fundamental in considering any fluvial geomorphological problem; look upstream and look downstream. In fluvial geomorphology the causes are seldom found in the same place as the symptoms.

In this case the river was historically channelized making it straighter and deeper. Although this planform was imposed on the environment, the system has adjusted to the slope, stream energy and sediment transport, in the years since, so that by and large the system was at an equilibrium, or at least geomorphogical change was gradual. Approximately ½ kilometre upstream of the reach in question a large section of river was restored in 2005 to its pre-channelised planform and the bed level raised.


Cartoon showing the long profile of the river pre-restoration (solid line) and the long profile after the restoration works (hashed line). The locally increased slope at the end of the restored reach acts as a focus for erosion.

The raised bed level and increase in sinuosity results in a decrease in bed slope (the river is now longer due to the increase in meanders). The decreased slope slows the passage of water through these restored reaches (as water flow is driven by gravity). However the presence of the bridge over the river downstream of our reach acts as a hard-point. The elevation and bed-level at the bridge remains fixed, the bed level here cannot be raised reliably and the river bed cannot erode due to concrete reinforcement. Therefore if we consider the long profile of the river there is now an increased gradient through our reach between the end of the elevated bed level in the restored section and the fixed elevation of the bridge. This increased slope has increased the speed of the water and the erosional power of the water (called shear stress). The following equation shows shear stress at the river bed is a product of acceleration due to gravity and water density (both fixed) along with bed slope and water depth. I won’t go into a fully mathematical argument of this point, but its sufficient to show bed slope is very important in erosion & transport of sediment.

Shear Stress

Shear Stress

Cartoon showing how tree roots can strengthen river banks; however down-cutting of the bed can result in lateral erosion below the depth of the roots

Cartoon showing how tree roots can strengthen river banks; however down-cutting of the bed can result in lateral erosion below the depth of the roots

As a result the bed in our reach has been eroded and down-cut, and a process known as “knick-point recession” has occurred whereby small steps in the bed profile erode back up the length of the river. The river bed is now sufficiently down-cut that it is lower than the depth of the roots of most of the riparian trees and thus at the outside of meander bends the banks are even more susceptible to erosion by the higher energy stream as the bank material is not held together and reinforced by root systems.

I hope this demonstrates that the “problem” here is not the weakness of the banks per se, but the newly increased power of the stream to do erosional work. Increasing the strength of the banks to arrest the erosion is never likely to work long term as the stream still possesses the same power/capacity to erode. Reinforcing part of the bank will often just concentrate force and erosion elsewhere (this same issue occurs in trying to protect soft coastal cliffs from erosion). In this case the erosion has been concentrated at the front end of the bank defences, eroding away the small upright posts first, then eroding behind and under the large logs eventually causing their failure.

Close up of erosion control failure

Close up of erosion failure

So what can be done to arrest this run away erosion? It’s not an easy problem to solve and highlights the problems of conducting reach-based river restoration in general. However going back to my analysis that we have an imbalance between stream power and bed/bank resistance, and given I’ve highlighted the issues of trying to reinforce the substrate the logical conclusion is to reduce the stream power.

There are three possible methods I can see as potentially working here.

1)      Remove and replace the bridge and attempt to re-profile the change in bed elevation along a much longer length of stream. This could be done by raising the bed through the reach in question and downstream of the bridge. By using detailed topographic data to calculate the energy of the re-profiled stream along with possible numerical modelling it should be possible to design a new long profile in which the shear stress of the flow is in equilibrium with the resistivity of the bed (its “critical” shear stress)

2)      Increase the sinuosity of the reach in question. By increasing the length of stream through the reach the distance along which the drop in elevation takes place is increased, thus lowering the bed slope, water velocity and shear stress at any given point. This is challenging as it is a lot of work, would be an imposed river planform (not “restoration”) and would be difficult to get right.

3)      Increase energy dissipation in the reach. The installation of large boulders and large pieces of dead wood in the channel (large woody debris) will increase turbulence and possibly induce steps (mini-waterfalls) all of which dissipate energy and thus lower erosion. I should point out what I am advocating here is not the sort of tiny pieces of wood which are typical of “large” woody debris installed in river restoration projects! This would require entire mature trees and very large boulders and would need to be done in the understanding that the objective is the decrease of overall erosion through energy dissipation, locally there may be an increase in erosion due to flow diversion and a lateral buffer zone would need to be established in which it was acceptable for the river to migrate.

If I was the river manager for this catchment I would advocate a pragmatic approach of combining 1 and 3. However it is important to acknowledge that in no way could this be described as “restoration”, these would be pragmatic solutions for erosion control. It is very attractive to argue that an even more pragmatic solution would be to establish a buffer zone for erosion, allow the river to get on with it and eventually establish a new equilibrium on its own. The remaining issue being to maintain the asset of the river crossing of some description.

The main point to emphasise here is that a well-meaning river restoration project, which has been well executed, has had unintended and undesirable consequences. Due to a general lack of monitoring of projects it is impossible to know how symptomatic of river restoration projects in general this example is; indeed I have no knowledge of whether this on-going erosion has been identified as linked to the restoration upstream by the river manager.

Due to practical and economic constraints it is often impossible to do anything but reach scale and patchwork restoration projects, but in these cases it is vital to project success to understand the geomorphological issues and to embed the project within the river system both upstream and downstream.

Posted in Uncategorized | Tagged , , , , , , , | 1 Comment

The rise of “river restoration”

I read Simon’s (a.k.a. @woodinrivers) great blog post on defining river restoration, and the differences between restoration, rehabilitation, re-creation and enhancement.

To build on this further, I thought it would be interesting to provide some brief statistics (via Scopus) on riv

er restoration in the academic literature, drawing a few conclusions and questions as I go. Take what you will from this rudimentary literature review!

There have been 1246 academic papers with “river restoration” or “stream restoration” in the title published between 1973 and today (July 2013). We can see the terms starting to gather widespread use during the 1990s, as the practice of river restoration took hold. There has been a visible increase in papers since the 2000s, coinciding with the Water Framework Directive, peaking in the late 2000s. The general increase in published papers, likewise, must be treated with caution – research output in most areas is increasing, but it is interesting to see the relative dates of the “rise” of this term in academic papers, compared to something like hydrology, which rose in prominence in the 1950s.

Number of papers on river restoration by year

Scopus results of academic papers published with “river restoration” or “stream restoration” in the title, shown by year, against those with “hydrology” in the title for reference.

There has been a substantial proportion of conference papers among those published. We might speculate that conferences have been an important way to share experience of river restoration – perhaps reflecting the role of practitioners in this field of research, who are more likely published in conference proceedings than in academic articles.

River restoration papers by type

Scopus results of academic papers published with “river restoration” or “stream restoration” in the title, shown by paper type.

The US dominates the research output in terms of number of published papers, and this is fairly typical in most academic research. Often the UK comes up as the next biggest publishing country, but I am interested to see the amount of work coming from China. There may be much we can learn from the experiences there in mitigating major abuses of river systems in the past.

River restoration papers by country of origin

Scopus results of academic papers published with “river restoration” or “stream restoration” in the title, shown by country of origin.

What are the subject areas involved in publishing research on river restoration? Predominantly they are the environmental and natural sciences, with engineering maintaining a substantial share. It is interesting to see that, despite the important human (social, cultural, political) dimensions of river restoration and water management, the social sciences are not represented in greater numbers. It is also of interest to note that under “other”, subject areas include medical sciences (papers assessing effects of river restoration and associated water quality on public health, disease, and wellbeing). We might suggest that the breadth of subject areas reflects the inter-disciplinary nature of river restoration.

River restoration papers by subject area

Scopus results of academic papers published with “river restoration” or “stream restoration” in the title, shown by subject area.

Google Trends shows search patterns since 2004, suggesting a declining interest in river restoration. Can we take anything from this? Perhaps we might propose that the “golden age” of river restoration has now passed. Physical works to re-meander or re-grade river banks have fallen out of favour, owing to costs and a growing realisation that restoring the physical appearance of a river does not always equate to improved ecology, flood risk management, or sustainability. The bigger challenge is “restoring” the processes that underpin natural functioning of rivers, and with that must come acceptance of the inherent uncertainty and unpredictability of natural river processes, making space for them against competing land uses (both urban and rural areas “squeeze” rivers and their riparian and floodplain habitats), and developing a fully integrated, long term approach to managing them.

—Edit note November 2013: graph showing number of papers by year changed to show alongside papers with “hydrology” in the title, for a useful illustration of the comparatively more recent rise of this field.

Posted in paper review, River Management, River restoration | Tagged , , | Leave a comment

What is River Restoration?

Picture1This is a post I’ve been meaning to write for a long time; What is River Restoration? I’ve started, abandoned, rewritten and gone off on tangents with it a dozen times as although it seems a simple question it touches on a number of issues such as why restoration happens (see previous blog post in this series) and which methods/techniques are most suitable. I was not able to find any definition on the River Restoration Website and the Wikipedia entry is decidedly vague, although does make useful distinctions with river engineering.

This is a very complex issue as it relates partly to context and partly to a whole lot of jargon and semantics. I’m going to explain first the context of why someone might consider river restoration and then uncritically explore the different sub-types of restoration as I see them, before trying to explain why this all matters.

Photo from Brooks, 2006 - Design guideline for the reintroduction of wood into Australian streams

Photo from Brooks, 2006 – Design guideline for the reintroduction
of wood into Australian streams

In its most basic form river restoration is taking a river channel which is deemed to be degraded in some way and attempting to improve the status of the river through some form of intervention such that the degradation is reversed or minimised. I have chosen a deliberately broad definition as there are a very broad range of projects under the umbrella of “river restoration” and my objective is to shed some light on the area whilst avoiding semantic debates. Indeed there can be a good case made that some projects that would in the past have been labelled river engineering are now talked of under the umbrella of river restoration as it sounds less invasive.

Broadly speaking river restoration is deemed necessary or worthwhile where there is a legacy of human river modification. This modification could be direct such as dredging, or indirect such as agricultural practices increasing sediment delivery, but generally the river in question has been altered in some way from its “pre-human” state. Previous river modification may have altered the processes within the river such that the way the river looks, behaves, or functions is not meeting societal expectations (e.g. local people, river uses, legislators, etc).

The term “restoration” implies moving something backwards towards a pre-existing state. You may think this means “river restoration” would involve assessing how the river used to look and function and recreating this; however in reality “river restoration” has become an industry term which encompasses virtually any work on a river which aims to improve something about the river. Perhaps “river restoration” should therefore properly be referred to as “river improvement”, but this raises a further semantic question; improvement from who’s point of view?

We can sub-divide types of “river restoration” to help shed some light on what is going on.


Photo from Brooks, 2006 – Design guideline for the reintroduction
of wood into Australian streams

The first type I’d define is actual “restoration”. This involves detailed assessments of how the river would have appeared and functioned at some imagined-Neolithic point in time; the objective being to recreate the river as it was before any human intervention. In practice this is usually impossible to achieve due to the difficulty in defining the target state of the river. Maps don’t go back far enough and although it is sometimes possible to recreate the geomorphological evolution of the river through modelling or sub-surface mapping, and the ecological composition of the environment through coring, such an exercise is usually beyond the scope of most projects. Furthermore it is exceptionally challenging to remodel a river in terms of its planform and ecosystem.

The second type I would define as “re-creation” this is where the target river has been subject to a clearly definable historic phase of degradation, typically short and largely geomorphological, for example channelization or installation of a weir. The restoration works aim to recreate the river morphology pre-degradation through study of historic maps and topography. This may or may not be linked to an attempt to improve the ecology of the river, however this type of restoration can occasionally be ‘for its own sake’; that is to say something has been done to the river, it is perceived as a historical mistake and thus an attempt to reverse it is undertaken to make the river more “natural”, the actual outcomes are almost secondary.


Photo from Brooks, 2006 – Design guideline for the reintroduction
of wood into Australian streams

The third type is “rehabilitation” where (usually) small scale processes or features are identified as absent or degraded in the target river; typically this will involve insufficient habitat for one or more species or groups of interest (e.g. salmon).The restoration works aim to improve the processes or features in question by manipulating the system in small way. This is probably the category with the widest applications and encompasses the vast majority of “river restoration” occurring around the world in terms of number of projects. Examples could include inserting dead wood into a channel to provide habitats, planting vegetation strips along the river bank to catch fine sediment laden runoff and adding gravels to a channel to raise the bed-level. Such works are highly local in effects, of fairly short term duration (perhaps just a few years at best), and are rarely monitored to determine success; indeed in some cases the actual desired outcomes are not clearly defined in the first instance. Implementation is often led by industry experts who have led many past projects and projects will often rely on such expert knowledge rather than detailed site assessment beforehand. The popularity of such projects is partly a result of their low cost, short led-time/preparation and they are a very pragmatic way of enhancing a river in the short/medium term.

As you can see this is largely an issue of semantics, but it is important to be aware that river restoration may not involve any actual restoring, but rather general improvements to a degraded river. All of the above types of project have the potential to be beneficial, and each has their advantages and drawbacks. Although the precise definition of river restoration can seem like an abstract semantic argument I think it is important to be aware of the subtleties. If stakeholders better understand what sort of restoration is being conducted at a site then their expectations of improvements to the river can be tempered. Problems will arise with public support of river restoration if people believe a project will deliver a SSSI graded wilderness stream where a scheme is just removing a weir!

In any specific river it could well be the case that full system restoration is unrealistic (maybe over time a forest has been replaced by a town), re-creation maybe insufficient and rehabilitation possibly unsustainable. In such a case if stakeholders want improvements to the river a pragmatic solution will be needed and expectations should be tempered by reality and clearly understood by everyone with an interest in the river.

In a future post in this loose series I intend to explore the arguments for and against catchment scale, process based restoration (full restoration above), often held up as an ideal in the scientific literature, but remaining relatively rare in practice.

Posted in Ecology, Geomorphology, River Management, River restoration, Water Framework Directive | 10 Comments

Uttarkhand floods in India

ImageIn response to the devastating floods in the Uttarakhand province of India there has been an almost inevitable shift from the initial shock and horror towards attempts to apportion blame for the events. In many ways this is something that happens in other natural disasters and mirrors theories on the 5 stages of grieving, where denial is replaced by anger and bargaining; in this context observers move from passively watching news footage of the developing disaster before talking either to their peers, or perhaps writing news articles or blogs asking “how could this happen?”, “what could have been done differently?” (Representing bargaining in the 5 stages of grief theory).

Whilst this process is completely natural and I would never think to criticise anyone’s emotional responses to a natural disaster, I think it is important to try and partition the different components of any disaster in order that we can understand what sort of things CAN be potentially controlled and what elements are to a degree inevitable. In that way energy (and anger) can be directed towards potentially minimising the impacts of future events. Many comments start initially talking about flooding before quickly moving on to other contentious topics like sustainable tourism, dams, ecosystems, etc which have next to nothing to do with people dying in floods.


Typical Uttarakhand provice uplands – very steep slopes and heavily forested. Photo courtesy of

From a hydrological point of view the first thing to understand is land use can change flood risk; for a fairly moderate rainfall event a catchment covered in grassland may experience a minor flood event, whereas one covered in forest may only experience slightly elevated, in-bank river flow (i.e. no flood). The forest intercepts more rain with leaves, rain filters into the ground quicker and thus generally gets to the river slower; minimising the potential for flooding. In urban areas concrete surfaces and drainage systems lead to very fast runoff and to rainfall getting into rivers very quickly, meaning flash flood risk is higher. However this effect of “complex” land cover such as forest, reducing flood risk compared to grassland/bare earth/concrete, decreases as we look at larger volumes of rainfall/larger flood events. Even in the most absorbent ground surface can only absorb a certain amount of water over a given time; therefore as rainfall rate increases larger volumes of water will not soak into the ground and will move quickly to a river as runoff instead. Generally speaking in the very biggest flood events land cover only makes a small impact on flood magnitude (although it can delay time to flood peak).


Aerial view of Sydney Harbour. This file is licensed under the Creative Commons Attribution-Share Alike 2.0 Generic license.

Looking at the Uttarakhand province it is around 86.6% forest cover (1), so from that respect has fairly ‘runoff resistant’ land cover. Increased urbanisation in the remaining 13% is unlikely to have much impact on the magnitude of very large floods. The event in question had a rainfall rate of 340-385mm in 24 hours (depending on sources). The annual rainfall is 1553mm/yr (1) and although the area experiences most of this during the months June-September due to the monsoon, this still represents ~25% of the annual rainfall in one day. The catchment area above the Maneri Bhali 1 dam (very close to the devastated town of Uttarkashi) is 4024 km2 (3). If we assume the rainfall rate was spatially uniform over the day across this catchment (unlikely to be true, but stick with me)  we can integrate that height of water over the catchment area to produce a water volume; in this case 1,488,880,000 m3. To put that into context that is about 3 times the water in Sydney harbour, or enough to fill 60,000 Olympic sized swimming pools. This figure is unlikely to be correct due to the way monsoon rain storms move, meaning rainfall is unlikely to be +340mm everywhere in the catchment during that day, some areas may have been largely dry, however this ‘back of the envelope calculation’ does demonstrate we are dealing with a lot of water over a pretty run-off resistant land-cover. In some crude ways there are similarities with UK events such as Lynmouth and Boscastle in that all these catchments are at high risk of flash-flood type events due to their steep uplands with impermeable geology (in the case of Uttarkhand quartzite with some tectonically derived secondary permeability due to fractures (2)).

The conclusion I’d draw from this is minor changes to land cover within the catchment are highly unlikely to have made much difference to the magnitude of the event, due to existing high percentage forest cover. Where bloggers and columnists are correct to say this was a “man-made disaster” is in flood risk mitigation. The keys in catchments like this are early warnings, evacuation/rescue plans and building controls.

Early warnings can take the form of meteorological triggers due to certain rainfall rates, or stream gauges upstream of urban locations to give a few hours’ notice of an impending flood wave. This needs to be coupled with local education to make sure warnings can get through and people know what to do on receiving them. These articles in Times of India  and The Hindu states that someone knew a flood was imminent in time to do something about it but no-one seemed to know how to get the message out, or what people on the ground were supposed to do with the knowledge.

Rescue plans need to include detailed information on where people will be evacuated to and general local geography.

One of the most important things, and where I assume things have gone badly wrong in Uttarkashi, is in controlling building on floodplains and close to the river. In my abstract academic bubble I’d like to advocate no building near rivers at risk of flash flooding, however I know that practically this is unrealistic. A more pragmatic approach would be to enforce strict building controls around the material foundations are set on (bedrock, not alluvium which can be washed away/undermined) and the strength/resistance of buildings to partial inundation.

Ironically the building of so many dams in the upper Ganga has received criticism, but may be the only way  governments could actively attempt to reduce or mitigate these types of flood. The creation of large impounding dams upstream of a vulnerable urban location could theoretically form a flood defence if the reservoir was left largely empty and just designed to store and slowly release a flood wave. Such schemes may exist somewhere, but I am not aware of any. The ecological implications of effectively building a concrete wall across a valley just to contain a rare flood event are tricky to say the least.

In this region the impacts of climate change on potentially increasing the magnitude and frequency of extreme weather events, as well as potentially glacial outburst floods from melting Himalayan glaciers, coupled with the risk earthquakes could cause a catastrophic failure of one of the many dams in the region means another flash flood disaster is a virtual certainty in this part of the world in at least the medium term. There is therefore an imperative for the risk to be recognised and for strategies to be developed to try and minimise the loss of life in such an event.

Some references:

(1) –

(2) –

(3) – Hydrology & Water resources of India (book)

Posted in Flooding, Hydrology, Meteorology, Politics | Tagged , , | 3 Comments