For the past 60 million years or more, the Earth's climate has been slowly cooling. Simultaneously, the great mountains of the Himalayas and the Tibetan plateau have grown. Recent theories have suggested that there could be feedback between these two events – that mountain growth causes climate change, which in turn causes more mountain growth. Writing in Nature1, Kelin Whipple of Massachusetts Institute of Technology, Cambridge, Massachusetts and colleagues put new limits on this idea.

Few dispute that the growth of high mountains alters climate by changing global wind patterns, causing weather systems such as the annual monsoon in India and Southeast Asia. A more controversial recent claim is that climate cooling may make mountains grow. The reasoning is that mountains are made of light rock which "floats" on the denser rocks below them. The average height of the mountains depends on the depth to which the light rocks below extend – the deeper the root, the higher the mountain. As mountains are worn away their average height gets lower, but only by about one-sixth of the thickness of the rock that is scoured away. This is because the mountain floats upward, compensating for the lost material just as a ship rises in the water as it is unloaded. But if river valleys are cut deeper into the rock, the height of the peaks above the average is greater so the mountaintops become higher. In turn, higher peaks alter global weather patterns more strongly, leading to increased cooling and so to greater erosion through increased rainfall, increased storminess and through the growth of glaciers.

Whipple and colleagues have analysed the effect of erosional forces on the relief of mountain regions. Starting from a well-documented relationship between the length of a stream and its potential to grind away its bed, they established a connection between the forces of erosion and the relief of the mountains. They come to the unexpected conclusion that higher rates of wearing away lead to less rugged mountains – that the height of ridges above valleys is reduced in times of high erosion, not increased. Although Whipple and colleagues could not carry out the same mathematical analysis for glaciers, they suggest that glaciation increases relief at most by a few hundred metres, and possibly not at all.

Whipple's conclusion is that increased erosion does not lead to mountain growth, except under exceptional conditions. So, if mountain-induced climate changes lead to faster erosion, then they cannot increase the rate at which mountains rise. In turn, the proposal that mountain-building and climate change can feed off each other is severely weakened. So, if we ask the question, which came first, mountains or climate change, then Whipple's work suggests that the answer is "mountains".