The WaterCentury

When we say we’re “running out”, what does that mean?

The planet isn’t an emptying tank

The phrase “running out of water” is useful because it gets attention. Arguably, it should. But taken literally, it gives the wrong picture.

Earth’s water continues to circulate through oceans, clouds, rivers, soils, ice and rock. It’s the water cycle, if you remember your KS2 science. Thankfully, humanity is not consuming the last water molecule and leaving the planet dry.

The problem is narrower than that, and more immediate. Water has to pass several tests before it becomes a dependable public resource.

It has to be fresh enough, clean enough, reachable, available at the right time and close enough to move without absurd cost. It also has to remain in rivers, wetlands and soils in sufficient quantities to support wildlife, land and food production. It’s a really big headache.

That is why the UK can flood in winter and still face severe droughts in summer. It is why a town can sit above an aquifer (the fancy word for an underground “sponge” made of rock) while nearby rivers run low.

Water scarcity is thousands of local systems losing balance in different ways, rather than one global tank emptying at a steady rate.

In practice, “running out” usually means the usable flow is no longer in the right place, clean enough, reachable enough, or reliable enough when people and ecosystems need it.

Where the Freshwater Actually Is

Freshwater can exist without being practically available

We know from the intro that about 2.5% of Earth’s water is freshwater (USGS). Even that figure can give a misleading sense of abundance.

Only a small share sits in lakes, rivers, wetlands, soils and the atmosphere in forms that are relatively easy to reach.

Groundwater is often treated as if it were one endlessly refillable underground lake. It isn’t. Aquifers recharge at different speeds depending on geology, soil, rainfall and the size of the system. Some refill over seasons. Others take decades, centuries or longer.

Pumping can therefore remove water far faster than the system replaces it, even while large volumes technically remain underground.

When water levels fall, the consequences aren’t limited to a lower number on a gauge. Springs can weaken. Rivers can lose the baseflow (the groundwater that helps keep rivers flowing between periods of rain) that carries them through dry weather. In coastal areas, saltwater can move into aquifers (Nature 2024), which is a serious problem for drinking-water supply.

And in heavily depleted systems of susceptible geology, the ground itself can compact, permanently reducing the space available to store water in the future.

Being underground does not make water renewable, reachable or safe to extract indefinitely.

We Take Water Faster Than Some Systems Recover

Agriculture, depletion and water-system debt

The pressure begins with the way people use the accessible remainder.

Agriculture accounts for roughly 72% of global freshwater withdrawals (FAO). Before you pick up your pitchforks, that does not automatically make all farming the villain. It does mean farming has to be central to any serious water strategy, and I’ll come back to that in Part Two.

On FAO estimates, irrigation allows roughly one-fifth of cultivated land to produce about two-fifths of the world’s food (FAO AQUASTAT). It is extraordinarily productive. But the local consequences depend on where crops are grown, how water is applied, what happens to the return flow and whether the source can recover.

Water withdrawn from a river and returned nearby is different from water evaporated, embedded in crops, polluted or pumped from an aquifer that recharges over generations.

Wasting water only compounds the pressure. A major Food and Agriculture Organization assessment estimated a blue-water footprint of about 250 cubic kilometres of surface water and groundwater for food produced but never eaten, using a 2007 baseline year (FAO 2013). For scale, using 2,500 cubic metres per Olympic-size pool, that is about 100 million pools. Or roughly 33 Loch Nesses (NatureScot).

A useful way to describe the result is what this project calls a local water debt. A basin, aquifer or ecosystem enters debt when withdrawals and damage repeatedly exceed replenishment and repair.

The phrase should not be treated as one global bank account. It describes specific systems being asked to provide more than they can sustainably replace.

Agriculture is not a side issue. The balance between crop demand, irrigation method, return flow, evaporation and recharge decides whether a source holds up or weakens beneath the field.

Scarcity Is About Place

There can be loads of water on Earth and still not enough where you live

Scarcity begins with unevenness.

Water arrives through weather systems, collects within river basins, seeps into aquifers and moves according to the landscape, not according to national wants or needs. Political borders, cities and intensive farming have been built on top of that geography, and they often fail to match it.

Around 2.1 billion people still lacked safely managed drinking water in 2024 (WHO/UNICEF). A 2016 model estimated that nearly 4 billion people experience severe water scarcity for at least one month each year (Mekonnen and Hoekstra 2016).

These figures describe different failures: one in safe and reliable service, the other in recurring periods when monthly demand exceeds available blue-water supply. Neither can be reduced to a single map of “dry countries”.

This mismatch can exist inside wealthy, wet nations too. One region may receive abundant rain while another holds the population, farmland or industry. Moving water is possible, but it requires pipes, pumping, storage, planning and public consent. Global totals can look comfortable while individual towns, farms and rivers face acute pressure.

The useful unit is the basin, the season and the supply system: where rain falls, where it is stored, who withdraws it, and what is left in the river afterwards.

Demand grows locally too. Extra homes, businesses and industry matter because they add connections inside a particular water-resource zone and distribution network. Ten thousand new homes do not matter because each person suddenly drinks dramatically more water. They matter because more taps, toilets, showers, gardens and workplaces are attached to a system built for a different pattern of demand.

Water Security Is Built

What happens after the rain falls?

Rainfall alone does not create water security. A dependable supply is built from catchments, soils, aquifers, reservoirs, treatment works, pipes, maintenance, regulation and long-term investment (UN-Water).

Two places can receive similar rainfall and still end up with very different runoff and infiltration, depending on what happens after the water touches the ground.

Compacted soil sheds rain quickly. Hedgerows, trees, permanent vegetation and better-structured soils can slow surface flow and create more opportunities for water to soak in (Environment Agency 2017).

The Pontbren trials in Wales measured far higher infiltration beneath tree shelterbelts (the fancy word for a line of trees and shrubs) than on nearby heavily grazed pasture: up to about 60 times higher in the studied plots.

That figure does not mean every hedge produces the same result. It shows that soil structure and vegetation can change how quickly rain leaves a field.

Before water reaches a reservoir or pipe, the ground has already made a decision for it: soak in, run off, evaporate, or carry sediment and pollution downstream.

After that, the problem changes shape. Water has to move from source to treatment, through trunk mains (the big pipes that carry bulk water), into service reservoirs, through local distribution pipes and finally to the tap. Each stage has a limit. A place can have water available at the source and still struggle if the network cannot physically move enough treated water quickly enough to where demand is rising.

And there is another problem here. The pipes themselves can only carry so much.

That is different from leakage. Leakage is water escaping before it reaches a customer. Conveyance capacity (how much water the pipes can physically carry) is the maximum amount the system can push through the network in the first place.

Not all reservoirs are lakes

When most of us hear “reservoir”, we picture a giant lake. But water companies also use covered service reservoirs. These store clean, treated drinking water closer to the people who will use it. You could live near one and have absolutely no idea it is there.

And a service reservoir can empty faster than the pipe feeding it can refill it. That can happen even while plenty of raw water exists somewhere else in the system. Suddenly “the reservoir is still full, why is there a hosepipe ban?” becomes a much more complicated question.

Climate Makes Timing Less Reliable

A very wet month can be followed by a very dry one

Climate change should not be described as every place becoming steadily drier.

The more useful warning is volatility: water arriving at the wrong speed, in the wrong season, or after dry weather has already lowered rivers and soils. Warmer conditions can alter rainfall patterns, increase evaporation, intensify downpours, and extend dry periods. Water can arrive in larger bursts and still become less dependable across the year.

2020 gave the UK a clear, documented example of a rapid wet-to-dry transition.

February was the wettest February on record for the UK: 209.1 millimetres of rain, equal to 237% of the long-term average (Met Office). At the time, it probably just felt like a miserably wet month. In hindsight, it became a useful example of how quickly the water picture can change.

Successive storms that winter flooded thousands of properties across England and Wales (Environment Agency 2020). Then the weather reversed. Spring became the UK’s sunniest on record, and May was England’s driest on record, with just 9.6 millimetres of rain (17% of average) (Met Office).

Some of that winter rain soaked into soils and aquifers. Some was held temporarily. Some moved quickly through rivers and drainage systems. Different catchments responded differently.

The main point is that record rainfall did not guarantee secure flows a few months later. Fast-responding rivers fell sharply as the dry weather continued (UKCEH). Receiving water and retaining it for later are different engineering and ecological problems.

More rain falling at once can still leave less water available later if it runs off faster than soils, aquifers, reservoirs, and wetlands can hold it.

The dashed river-response line is there to explain the general pattern described by UKCEH. It is not measured river-flow data.

The rainfall figures all use the same UK-wide Met Office HadUK-Grid series. The current dataset gives February 2020 as 213.7mm. The Met Office’s original announcement at the time used 209.1mm before later revisions to the dataset. The 9.6mm May figure in the text is for England alone, which is where the record sits; UK-wide, and so on the chart, May 2020 came in at 32.8mm.

One year cannot prove an entire climate trend. It is simply a very clear example of how quickly conditions can change.

Water Scarcity Spreads

When water runs short, other things start breaking

Water rarely creates a crisis on its own. It amplifies other pressures. When it runs short, farms can lose output, ecosystems lose flow, power stations may lose cooling water, freight routes become unreliable, and households face higher costs.

The consequences can travel far beyond the reservoir, aquifer, or river where the shortage begins. The European drought of 2022 showed that cascade clearly.

Almost two-thirds of Europe’s rivers ran below average against the 1991 to 2020 baseline. 26% were classed as exceptionally low (Copernicus). Forecast yields fell sharply for maize, soybeans, and sunflowers. Expectations were cut to roughly 12 to 16% below the five-year average (JRC).

Low river levels restricted freight on the Rhine. High water temperatures and limited flows constrained electricity generation in parts of Europe (IEA).

That drought unfolded during a wider energy crisis, so it would be wrong to blame water for every price increase that followed. The narrower lesson is still significant. One period of water stress coincided with, and contributed to, pressure on food production, energy generation, and transport at the same time. Low flow reduced river carrying capacity, heat constrained cooling, and irrigation restrictions arrived while crops were already under stress (UNCCD 2024).

Water stress spreads through the parts of the economy that need flow, cooling, transport, irrigation, or clean supply at the same time.

When it isn’t, we find out very quickly.

So What Can Actually Be Changed?

Quite a lot, as it turns out

There isn’t one universal answer. The evidence points to a stack of fixes. In some systems, fixing leaks, cutting pollution or reducing waste is simply cheaper than developing new supply.

Better irrigation, healthier soils, restored wetlands and reducing demand can all lower the pressure. It makes sense to do these before we start building new reservoirs.

Britain has no excuse to treat this as a distant problem. The Environment Agency’s 2025 national framework paints a stark picture. Under their “Do Nothing” scenario, England could face a public-supply shortfall of five billion litres a day by 2055. The wider economy needs another billion litres a day on top of that (Environment Agency 2025).

Leakage reduction and lower demand won’t be enough to close the gap. That is why current plans include new reservoirs, recycling schemes and desalination (Environment Agency 2025).

The contrast between household restrictions and systemic loss is politically poisonous. Ofwat’s figures for 2024 to 2025 show about 2.87 billion litres of water leaking from company networks in England and Wales every single day. Their headline measure uses a three-year average, which sat at roughly 2.9 billion litres a day between 2022 and 2025. In water-industry units, that is 2,869 million litres a day (Ml/day) (Ofwat).

We need more storage. We also need functioning pipes, protected catchments and abstraction rules that actually protect rivers. And we need maintenance budgets that survive beyond the next billing cycle.

Desalination and water reuse might help too. But they come with their own headaches. They are energy-intensive, and involve complex logistics like brine disposal and marine intake.

I am not trying to choose between these options here. I am just showing why we have to make a choice. We will get into the specific trade-offs in the next part.

The goal is simple: resilience. We need enough demand reduction, leakage control, storage, reuse and catchment repair so that a dry year does not become a national emergency.

Unfortunately, I have mostly discovered that there are about fifteen answers and they all have problems. Excellent.

The Next Question

From the problem to the investigation

We now have the shape of the problem. Usable supply is limited, unevenly distributed and exposed to depleted aquifers, damaged landscapes, ageing infrastructure and more volatile weather.

That is a serious problem, not a prophecy. We’re not doomed. Some pressure can be relieved by fixing leaks, retaining more rainfall, restoring catchments, modernising agriculture, reusing water and building storage.

Other shortages may require additional supply. All major options carry trade-offs. Sadly for me, “technical potential” means very little if a proposal cannot survive environmental scrutiny, public budgets or the soul-crushing state of political reality.

So the next part will not begin with a preferred answer. It will compare the available levers: how much water they save, what they cost, how quickly they scale, where they work, and which problems they shift elsewhere. Only after that investigation will I propose a solution.

Then we have to deal with politics.