◈onway

Small water · the wisdom we paved over

Ten thousand tanks, not one dam.

Long before big concrete, undivided India was covered in small built water — tanks and lakes, stepwells and earthen check-dams — that quietly survived both of nature's extremes. Then, in about sixty years, we broke it. The way back is not bigger. It is smaller, and everywhere.

Under the Mughals in the north, and the kingdoms and — above all — the village communities before and alongside them across the whole subcontinent, water was built into the land. Reservoirs and pleasure lakes, the deep cool baoli stepwells, the temple tanks, the canals that Shah Jahan's engineers revived to carry the Yamuna to Delhi. And beneath the grand works, the real backbone: countless small community tanks and check-dams, each with its own name in its own tongue — the eri of the Tamil country, the kere of Karnataka, the ahar-pyne of Bihar, the johad, tanka, khadin and bawdi of the dry west.

What made them wise was not their size but their number and their placement. Spread in their thousands across a watershed, high on the slopes as much as down in the valleys, they did the one thing a single great dam cannot: they met both extremes at once. In the monsoon they caught the runoff before it could gather into a flood, shaving the peak and holding the soil. In the dry months that same water, sunk slowly into the ground, came back up in the wells. A land dotted with small holds does not swing between drowning and thirst. It stays even.

THE OLD WAY — many small holds slow · spread · sink → recharge flood peak shaved · water kept for the dry months THE LAST 60 YEARS — one straight rush one dam narrowed · straightened · built over all of it downstream at once → flood, then drought
The same rain, two catchments. Many small holds high in the watershed slow the water, sink it, and even out the year; a single dam below a straightened, built-over stream concentrates everything into one flood and leaves little for the dry season. Size was never the strength — number and placement were.

How we broke it, in sixty years

From about the 1960s the small water fell out of favour. The age chased scale — the big dam, the deep tubewell, the canal from far away — and the thousands of little tanks, no longer needed on paper, were left to silt up and forgotten. Worse, the ground they sat on became valuable. Tank beds were filled and built over; floodplains were sold; and streams — the veins of the whole system — were narrowed, straightened, culverted, and in places simply blocked. A city that buries its tanks and chokes its streams gets the exact double punishment the old system was built to prevent: it floods in the rain, because the water has nowhere to spread, and it runs dry after, because nothing was left to sink.

The proof that small still works

This is not nostalgia; it has been tested, recently, at scale-by-numbers rather than scale-by-size. In the dry Alwar district of Rajasthan, from the mid-1980s, communities led by Rajendra Singh and the Tarun Bharat Sangh rebuilt the old johads by hand — one small earthen crescent at a time.

8,600+
johads rebuilt across 1,086 villages, by hand
100 m → 3–13 m
the water table rose back toward the surface
5 rivers
seasonal rivers, incl. the Arvari, began to flow year-round again

No single structure did this. Thousands of small ones did — each too small to matter alone, and together enough to bring a river back from the dead.

So: small, and everywhere — not big, and central

The instinct this whole site keeps returning to holds here too: meet the water on its own terms. Modern watershed engineering has the tools to do the old work better — contour bunds and swales, recharge pits and check-dams, restored tanks, and, most of all, streams and floodplains left open to do their job. What matters is the scale. The answer is not one more enormous project. It is ten thousand small ones, distributed across the catchment, sized to a village and maintained by it — the pattern that a wave of thinking in the last century, permaculture among it, rediscovered and named, but which the tanks had known all along.

Build small water. Unbury the tanks. Let the streams be streams. It is cheaper than the disasters it prevents, and it is the kind of work that, like the old refuges, keeps giving for a hundred years.

Where in the world — the same idea, drawn for each dryland

This is not India's alone. Almost every dry or degraded land has its own version, and the design is the same move made local: match the structure to the slope, the soil and the rain; put it high in the catchment; and keep it small enough for a village to build and maintain. Below are real drylands that need it — each pinned to a proven precedent nearby, so none of this is theory.

Pits & bunds — sheet-flow drylands Check dams & terraces — eroding slopes Sand & underground dams — seasonal rivers Recharge & storage — tanks, qanats, amunas

Design proposals matched to real, working precedents — not funded projects. Any real scheme must be led by local hydrology and the community that will keep it. Tap a marker for the designed system and its precedent.

Water levels & small ponds

Why a small pond moves the water table

A pond, tank or check-dam does more than hold water on the surface. By keeping the rain in place for hours or days it lets the water soak down — recharging the aquifer and lifting the level in every well nearby. In Alwar, 8,600 small johads raised the water table from about 100 m deep to 3–13 m and brought dead rivers back to flowing. The same ponds shave the flood, because water held is water not rushing downstream. Small ponds are the one lever that moves both the shortage and the flood at once.

Choose an area — its water future, and what could be done

Type a place, a country, or a word like “drought”, “flood” or “aquifer” — or use a shortcut. Almost everywhere now swings between too little and too much. For each area, two answers: what a year of willing hands could build, and what a government could do ahead of the crisis instead of rationing and paying for the flood afterward.

~4 billion → ~60% by 2050
people facing water stress at least one month a year — now, and by mid-century (WRI Aqueduct).
$1 ahead ≈ $6–$10 back
return on getting in front of disaster vs. paying after — up to about $7-to-$1 for river floods (Pew · Swiss Re).
Or browse every area (by region) — choose from the full list

Live water data that exists to tap (most needs a small server-side proxy — no direct browser access): USGS Water Services (US streamflow & groundwater) · USBR RISE (reservoirs — Mead/Powell) · WRI Aqueduct (global risk) · NASA GRACE (groundwater from space).

Before you need it — your shelter, and who to call

The best time to learn where you'd go and who you'd call is a calm one. Type an address or town — home, school, or campus — and we'll show who to reach and how to find your nearest shelter. Do it once, in peaceful times, and it's there when it counts — for kids, new arrivals, and lifelong residents alike.

Candidate check-dam sites

Grounding & further reading: CSE / Down To Earth — “Dying Wisdom” · Rajendra Singh & the johad revival · India's traditional water systems. Figures as widely reported; treat as illustrative of the pattern, not precise to the litre.

← Floods Fire & mudslide Higher Ground The letter on limits