The Himalayan mountain range is high but has deep roots that provide isostatic compensation. But the Himalayas are only the southern rampart of a vast plateau - Tibet. Both have the same average elevation of just under 5 km above sea level. The difference is that the Himalayas have more extremes (high mountains and deep valleys) while Tibet is smoother. Both, and the other mountain ranges of central Asia, are the surface manifestation of thickened continental crust. This thickening presumably happened after India collided with Eurasia and continued to move north. The plate tectonic data suggest that between 2000 and 2500 km of convergence has happened since the continents first came together. But how is this northward motion accommodated?

A central concept in understanding collision mountain ranges is the idea of a "suture" - or join - between the two continents. It is generally agreed that the Himalayas are formed by crustal-scale thrusting. This stacks up panels of crust. One of the first thrust models - proposed 40 years before the formal development of plate tectonics - was made by Emile Argand, the brilliant Swiss geologist. Critically he suggested that India was pushed beneath Eurasia - creating the Tibetan plateau. Another model, proposed in the 1980s by Philip England and Greg Houseman, is that the Eurasian side of collision thickened up, with relatively little stacking on the Indian side. Yet another model, advocated by Paul Tapponnier since the 1970s, is that Eurasia is being squirted eastwards and India acts a near-rigid indenting ram.

We can test between these models. First off, we can investigate if deformation is concentrated to the north of the suture (Asian side, as the England-Houseman and Tapponnier models both suggest) or on the south (Indian side, as required by the Argand model). Click here for setting up the test. The key information comes from information on the palaeolatitude of the southern edge of the Tibetan crust (a chunk of crust called the Lhasa block, named after the Tibetan capital), which comes from palaeomagnetic studies. Click here for these results. What do the data imply?

If most of the convergence is accommodated on the Asian side we now have to choose between the two models of thickening vs squirting (extrusion). We can use the earthquake data and the topography. Interestingly these give opposite results. The seismicity show the Tibetan crust not to be thickening but thinning in an east-west direction. So this looks like it's being squirted sideways. But this doesn't explain the topographic elevation of the Tibetan plateau - and hence its great crustal thickness. Perhaps a compromise is needed - with the crust thickening up at first but now it is spreading out sideways. Even Argand was partly right, because the Himalayas are telling us that the Indian crust is being stacked up. All this shows how complicated orogeny can be. Most critically - the end result is that deformation in the continents is far most complex than in the oceans. Plate boundaries are marked in the continents are broad areas of strained crust (picked out by the wide tracts of seismicity) while the oceanic plate boundaries are narrow (as seen in the earthquake map)
We can see how rocks respond to all this orogeny by visiting outcrops on a natural section through the Pakistan Himalayas - up the Karakoram Highway.

Himalayas front page