Drive out of Delhi towards the airport and the ground lifts, briefly, into a line of broken quartzite. In the city the Aravalli ridge stands only a few tens of metres above the plain, and for most of the year it is a scrubby green band behind the highway. It is also the northern remnant of a mountain system raised by the collision of two cratons, possibly once as high as the Himalaya, and wearing down for something close to a billion and a half years.

Almost everything written about this range begins with the word barrier and stops there. The interesting part is the mechanism. A wall tens of metres high should not stop a desert. It does, and the reason is not height at all.

A mountain that has been spent

Geologists place the Aravalli–Delhi belt among the oldest fold systems on earth. The rocks belong to two stacked successions — the Aravalli supergroup below and the Delhi supergroup above — folded and metamorphosed in the Proterozoic when the Bundelkhand and Marwar blocks came together. What survives today is not the mountain. It is the mountain's resistance: the quartzites held, the softer beds around them did not. A ridge here is a skeleton showing through.

Set the dimensions against that age and they look almost comic. Six hundred and ninety-two kilometres from the Delhi ridge to Palanpur in Gujarat, up to a hundred kilometres wide in Rajasthan, reaching 1,722 metres at Guru Shikhar on Mount Abu — the highest ground between the Himalaya and the Nilgiris, which says more about the surrounding plains than about the peak. A young range of that height would be a serious obstacle. This one has been sanded down to ankle height in the north.

Ankle height turns out to be enough, because of what it is standing in front of.

What a low wall actually does

Wind-blown sand does not fly. In the Thar it moves by saltation: grains hop along the ground within the lowest metre or two of the atmosphere, each landing launching the next. A crest two or three hundred metres high interrupts that process in the crudest possible way — the ground goes up, and the sand has to climb. Most of it does not. What gets over is what the wind can lift higher, and that fraction is what the eastern plains receive.

West to east across the Aravallis: sand moving by saltation is stopped by a continuous crest and passes through a breach CROSS-SECTION, WEST TO EAST Thar crest Indo-Gangetic plain sand, hopping in the lowest metre breach the fraction that gets through + rain The crest is what does the work. Height helps only as far as it keeps the line unbroken. Every breach is a road for sand, and the north end has most of them. Schematic section. Vertical relief exaggerated; not to scale.
A continuous crest breaks the hop; a breach returns the whole downwind plain to the sand's path. The ridge's value is a line, not a summit.

The second job is water. The range is a divide in the plainest sense: the Luni and its tributaries run west and south-west off the western flank and end in the Rann of Kutch, while the Banas and its feeders run east into the Chambal and so to the Yamuna and the Bay of Bengal. One low ridge separates two different seas. Every river that rises on it is water that a farmer two hundred kilometres away either gets or does not.

The third job is invisible from the road. Fractured quartzite is a decent aquifer, and the ridge is the spine along which rain enters the ground for Jaipur, Alwar and southern Haryana. Wells at the foot of the hills are not drawing on the plain; they are drawing on the hills. Cut the ridge, and the recharge that fed them goes somewhere else.

The gap is the story

In the south the Aravallis behave like a mountain range: continuous, high, largely under protected forest, with Sariska and Kumbhalgarh and the Abu massif holding the line. North of Jaipur they come apart. Alwar, then the broken country around Gurugram and Faridabad, then the Delhi Ridge, which ends under alluvium near the Yamuna. That fragmentation is natural — an old range dies by dissolving into hills. What is not natural is the speed at which the gaps have widened.

The range is a mineral belt, and that is the whole problem in one sentence. Makrana marble, Khetri copper, Zawar zinc and lead, plus granite and building stone lying at the edge of the largest construction market in the country. The same quartzite that makes the ridge useful as a barrier makes it useful as a quarry. There is no version of this economy in which the two uses do not compete for the same acre.

Mining here has been in court for four decades. The M.C. Mehta litigation produced a series of orders restricting quarrying in the Haryana and Rajasthan stretches, including directions in 2002 and 2009 and an order in 2018 to remove illegal structures. The existence of that litigation is itself a finding: the prohibition had to be repeated because the practice continued. One widely cited assessment puts the share of mining leases holding environmental clearance at a small fraction of the total, and estimates that the central Aravallis have lost about a third of their forest since the mid-1970s. Remote-sensing numbers of this kind move with the definition of a forest class, and should be read as a direction rather than a ledger.

There is a further wrinkle that matters more than it looks. Nobody agrees on where the Aravallis end. A range that has eroded into scattered hills can be defined by geology, by elevation, by contiguity, or by what has already been notified — and each definition puts a different set of plots inside or outside the law. Arguments about the boundary sound like pedantry and are not. They decide who may quarry.

Why the north is the vulnerable end

The asymmetry is easy to miss because the south is the scenic part. In the south the ridge is high, continuous, sparsely settled and already largely protected. In the north it is low, broken, and standing next to an urban region of thirty million people that needs stone, land and road alignments. The northern end is also the end the dust corridor points at: the pre-monsoon winds come from the west, and whatever the ridge fails to stop arrives over the same plain that grows the country's wheat.

So the same feature is doing two jobs with opposite requirements. As a barrier it needs to be continuous and left alone. As a mineral belt and a land bank it needs to be opened. The south can satisfy both, because nobody is building there. The north cannot, and the north is where the barrier matters most.

Planting a belt is not rebuilding a wall

The current answer is a green wall: a belt roughly 1,400 kilometres long and five kilometres wide, announced in 2025 and starting in five Haryana districts, with a stated target of about 1.1 million hectares under restoration. The logic is sound as far as it goes — native species, windbreaks, revived water bodies, dust interception. A belt of trees does slow surface wind, and slowing wind is how you stop sand from hopping.

The limit is geometric. Trees are points; a barrier is a line. A five-kilometre belt planted on either side of an operating quarry does not close the quarry. And planting in a 500-millimetre rainfall belt with a falling water table is not free — saplings need watering for years, in a district where the aquifer is already the thing being defended.

The unglamorous measure is the one that would actually work: settle the boundary, notify it, and refuse clearance in the gaps. That is administrative work with no photograph in it, which is roughly why it keeps losing to planting drives. But a barrier is not restored by adding trees to the parts that are still standing. It is restored by closing the parts that are not.

What is left of a worn mountain is not height. It is position — and position is worth something only while the line holds.

Sources and checks

  1. Geological Survey of India Aravalli–Delhi fold belt: stratigraphy and structure of the Aravalli and Delhi supergroups
  2. Supreme Court of India M.C. Mehta v. Union of India — orders on mining and construction in the Aravalli hills, 1991 to 2018
  3. Forest Survey of India India State of Forest Report, state chapters for Rajasthan, Haryana and Delhi
  4. Centre for Science and Environment Analyses of Aravalli forest cover, mining leases and clearances, 2025
  5. Ministry of Environment, Forest and Climate Change Aravalli Green Wall project documents and plantation targets, 2025
  6. India Meteorological Department / IITM Studies of dust transport and near-surface wind over the National Capital Region

Last checked 1 November 2026 Yellowsea Huang Corrections