It took me a while to get my head around the concepts in this post, so bear with me. This is aimed especially at eNGOs and other campaigners for chalk streams, because the more people there are who understand this counter-intuitve idea, the better. 

Here it is: you can save many chalk streams from unsustainable abstraction by conceivably using the aquifer in times of low flows and drought.

That is a head-muddler. But this idea could unlock real abstraction reduction, making the bad much better in the foreseeable future. This is far, far preferable in my view than holding out for a perfection (natural aquifers) that will never come.

It starts with my best attempt at explaining what I understand of the complexities of the interactions between groundwater, river flow and groundwater abstraction. Given that I vainly spent a long night in a hut in Iceland trying to explain the very same ideas to two angling friends of mine (they were belligerently uncomprehending in a (successful) effort to annoy me), this will be no easy task.

It is complex … kind of. It’s also quite simple really. Rather as the moon affects the tides, a simple idea leads to a complex set of manifestations.

Idea 1. Chalk streams flow from underground.

If you’re reading this blog you’ll already know that chalk streams derive most of their flow from groundwater. Rain sinks into the ground filling fractures in the underlying chalk and then lower down the slope it seeps out again as springs to create a chalk stream.

Idea 2. The level of the groundwater drives the flow in the river.

This is pretty simple. I used the bucket analogy before. Drill a single hole in the base of a bucket. Fill the bucket with water. As the bucket fills gravity drives water at an increasing velocity out of the hole. Now stop filling and let it empty. The flow diminishes to a trickle. EVERYONE gets this because it’s the same when you pee!

Idea 3. Groundwater rises in winter and falls in summer.

If you pour water into the bucket faster than water can leave it through the hole(s), the level in the bucket rises. If you stop pouring water in, the level falls as the bucket drains. This is exactly the same with a chalk aquifer. In winter, when it rains a lot, and it’s cold and the ground is wet and nothing is growing, more rain flows into the aquifer than can leave it and so the groundwater level rises. In summer, much less rain – if any – reaches the aquifer and so the groundwater level falls.

Idea 4. The higher the groundwater rises up the valley, the more the water pours out of it.

As groundwater level rises, stream flow increases. But not in a linear way as it would with a single hole at the base of a columnar bucket. In fact for every unit of rise in groundwater level, flow will increase by approximately X^{2 to 2.5} . Kind of like having twice as many holes at each level in the bucket as the level below.

There are a number of reasons for this which were debated at a recent groundwater conference. There is a summary of these ideas in Section 2 of John Lawson’s report Flow Recovery Following Abstraction Reduction which we updated following the conference and contributions from the likes of Rob Soley and Alessandro Marsili.

In short, this non-linear response is probably caused by a combination of: 

• the shape of the valley – if you imagine the groundwater filling the valley bottom and hillsides, assuming a perfect V- shape valley, for every unit increase the groundwater rises the area of saturated zone exposing springs rises two-and-a-half fold. Chalk valleys are not quite V-shaped but that’s the general idea.

• the fracture density in the chalk – which increases in the valley bottoms and with altitude. At depth chalk is very solid, but in the valley bottoms and higher up the slope and where water has flowed for thousands of years, the fracture density is much greater and the flow pathways are bigger.

• layering within the chalk – chalk accreted in layers under varying climatic / geological conditions and these layers are in turn interrupted by bands of clay and flint. These layers and the varying permeability and transmissivity can influence the way groundwater reaches with the surface.

• as the surface flow pathways lengthen (winterbournes rising higher and higher up the valley) the groundwater pathways shorten.

The fracture density and layering in the chalk, the shape of the valley and the length of flow pathways, all conspire to mean that when chalk valleys fill, flows will rise exponentially.

Idea 5. The impact of a constant groundwater abstraction has a varying impact on varying flows through the year

This is where things gets a bit more discombobulating. All of the above essentially means that as groundwater rises, flows increase exponentially. If that is true, then the reverse is true. For every unit of decrease in groundwater level, flows decrease exponentially.

This means …. drum roll … groundwater abstraction (which lowers groundwater levels) has a greater impact on high flows than low flows! This is a totally skull-tightening idea. Everyone thinks the reverse must be true. But it isn’t.

Groundwater levels and groundwater abstraction

Let’s start with the impact of groundwater abstraction on groundwater levels. In a natural aquifer system, the discharge from the valley must equal the recharge over time. Natural recharge = natural discharge / Time. This stands to reason: if it didn’t the valley would either fill to overflowing or empty (because over time one would exceed the other). 

Natural recharge derives from rain and natural discharge from river flow (and some evapotranspiration and flow through the ground). If I add another form of discharge in the form of abstraction, then the former natural discharge MUST go down. If it didn’t, the aquifer would progressively empty until there was no water left (an aside … hydrogeological literature generally describes anything less than draining the aquifer “sustainable”, because the aquifer is being lowered to a new dynamic balance, not mined. This is not the same as ecologically sustainable, however).

Look at it as simple numbers.

Natural recharge (10) = natural discharge (10) / Time.

Natural recharge (10) = abstraction (5) + natural discharge ? / Time.

What’s the new natural discharge? 5, obviously.

Now, as I showed with the bucket, the ONLY way in which the former natural discharge can go down is through a reduction in groundwater levels. If groundwater levels didn’t go down, then because the discharge is driven by the groundwater level the natural discharge would remain the same. As shown above, that is impossible.

Theis, the Isaac Newton of groundwater theory, wrote all this in 1940. The only way that the former natural discharge can go down (and balance the equation) he wrote, is by a reduction in the “thickness of the aquifer”. 

Okay, so pause and get your head round all that. 

• a single unit of rise or fall in groundwater level has a (very roughly) two-and-a-half fold impact on flows. 

• groundwater abstraction lowers groundwater levels.

• ipso facto a single unit of reduction in groundwater level at high groundwater levels has a much greater impact on flows than a single unit of reduction in groundwater level at low groundwater levels.

It still hurts the head, but the discombobulating stuff above means that at high groundwater levels groundwater abstraction reduces flows by quite a lot more than 100% of the amount abstracted. And conversely, at low groundwater levels groundwater abstraction reduces flows by quite a lot less than 100% of the amount abstracted. Albeit over time groundwater abstraction must reduce flows by (essentially) 100% of the amount abstracted (it’s generally less than that for reasons that aren’t that important to the general concept, but basically because not all discharge occurs in the form of flow).

See the chart below to see what the Chalk Streams First modelling indicates % flow recovery would be if abstraction was reduced to zero in the River Ver. It varies through the flow cycle.

Idea 6. Groundwater abstraction at low flows is like a credit card.

The obvious question is … if groundwater abstraction at low flows reduces those flows by a lot less than 100% of the amount being abstracted, where the bloody hell is the rest of the water coming from? The answer: if it’s not a direct reduction from flows at the time, it is coming from aquifer storage.

This is easy to understand if you think of a large abstraction next to a small and diminishing stream. In the winter when the stream is gushing, there is more than enough water to satisfy the pumping. In the summer the stream reduces to a trickle or perhaps even dries up. But the pumping continues. At this point the abstraction is clearly not taking water from stream flow because there isn’t any. Another aside … I’ve read hydrogeologists describe this state as abstraction having “no further effect on flows”. This might be literally correct at the time. But it is misleading. The abstraction is effecting future flows. 

At times of low flow and into droughts, groundwater abstraction increasingly draws on storage, upon which future flows are built. If you unnaturally drain the aquifer, it will clearly take longer to fill when it starts raining again, all before the flows in the river can respond to the rise in groundwater levels.

Therefore groundwater abstraction at low flows is like a credit card: much more a debt against future flows than an impact on present flows. This is a key idea behind the confusing concept of using groundwater abstraction to unlock abstraction reduction .

Idea 7. If you turn off the pumps you get greater flow recovery at high flows than low flows.

Essentially what all this means is that when you cease or lower abstraction you get well over 100% of the amount no longer abstracted at high flows and much less than 100% at low flows. That is what the chart above shows on the River Ver.

And this is the Achilles Heel of the Chalk Streams First idea. 

Water resources needs a constant supply of water. Groundwater abstraction gives that. Chalk Streams First says “turn off (or down) the pumps and take the water from river flows much lower down the catchment”. And while you get loads of water back in winter, you get less back in summer. Generally, you must have a storage reservoir to make it work and balance out the varying recovery rates into a constant and reliable supply. 

John Lawson – who came up with the Chalk Streams First idea – has long known this. We argue (with empirical evidence) that the flow recovery at low flows is actually much higher than the most pessimistic predictions claim, but nevertheless this variation in response is an issue we have to address. The answer is a reservoir.

BUT … then you get to the prolonged droughts when water companies are under real pressure. In these times, the flow recovery could conceivably drop even lower. What to do? The public must have water. This low flow recovery at very low flows in long droughts threatens the whole idea of reducing abstraction through schemes like Chalk Streams First. Especially now that we have to plan according to 1:500 year contingencies.

Idea 7. In droughts use groundwater abstraction to guarantee public water supply … so long as you’ve turned the abstraction right down to ecologically sustainable levels 95% of the time.

The insurance against the Achilles Heel of low flow recovery in a drought is a groundwater-fed public water supply scheme. There is one in existence already called the West Berkshire Groundwater Scheme (WBGWS). It is a series of wells in the Berkshire chalk that can, in extremis, be turned on and deliver a large amount of aquifer water into the Berkshire chalk streams, from where it flows to the Thames to be captured into the London reservoirs. The scheme is used very, very rarely: no more than once every 25 years. But it’s there. And it guarantees water in a drought.

The impacts on the chalk streams are a) one of flow relief in the drought, because the flows get boosted. Albeit – and I have to emphasise this – flow augmentation in not the aim of the scheme, it is a bi-product. And b) at the end of the drought, when the pumps are turned off, the aquifer must recover before flows return to natural levels, so you get lower flows the following year.

But this is crucial: in modelled scenarios, the flows in the year of recovery are still better than they would be if abstraction ran all the time as happens at the moment in streams like the Ver, Misbourne and Beane.

So WBGWS type schemes could unlock Chalk Streams First type abstraction reduction in other settings, such as on the chalk streams of the Colne, Lea and Ouse (even the Darent). As such a scheme would insure against the public supply deficit in droughts created by replacing upper catchment groundwater abstraction with lower catchment surface water abstraction (the Chalk Streams First concept).

BUT …the Environment Agency is very cautious of such schemes

This is understandable because there have been some bad schemes in the past. But flow augmentation to compensate for the collateral damage of abstraction is a different thing altogether. 

Some schemes were developed in the past whereby to compensate for abstraction (which had dried the stream) water was pumped from the aquifer into a losing reach of stream and the whole thing was a highway to nowhere.

Other times the concept of augmentation is used to justify continuing, unsustainable abstraction. These schemes have given the whole idea of flow augmentation a bad rap, and one that has stuck like glue.

BUT, I would argue that we need to be more pragmatic and open minded than a presumption against these schemes if we are to achieve the heretofore irreconcilable goals of reliable public water supply and restored chalk streams. Aquifers in the south east are managed one way or another. We need to make sure they are managed mindfully to achieve the specific outcomes we want and in this regard holding out for “natural” when a more flexible approach would unstick hopeful schemes such as Chalk Streams First would surely be counter-productive?

I understand the Environment Agency may be consulting on this topic later in the year. I know from many discussions I have had with chalk stream advocates that the ideas I have outlined above will be surprising and counter-intuitive to most of us, as indeed they are to me.

But it is vital we give the EA the encouragement to take a flexible, if ultra cautious approach, because the gains of doing so could be massive.