Petrology of Indus River sands: a key to interpret erosion history of the Western Himalayan Syntaxis

The Indus River has been progressively transformed in the last decades into a tightly regulated system of dams and channels, to produce food and energy for the rapidly growing population of Pakistan. Nevertheless, Indus River sands as far as the delta largely retain their distinct feldspar- and amphibole-rich composition, which is unique with respect to all other major rivers draining the Alpine-Himalayan belt except for the Brahmaputra. Both the Indus and Brahmaputra Rivers flow for half of their course along the India-Asia suture zone, and receive major contributions from both Asian active-margin batholiths and upper-amphibolite-facies domes rapidly exhumed at the Western and Eastern Himalayan syntaxes. Composition of Indus sands changes repeatedly and markedly in Ladakh and Baltistan, indicating overwhelming sediment flux from each successive tributary as the syntaxis is approached. Provenance estimates based on our integrated petrographic-mineralogical data set indicate that active-margin units (Karakorum and Transhimalayan arcs) provide similar to81% of the 250 +/- 50 10(6) t of sediments reaching the Tarbela reservoir each year. Partitioning of such flux among tributaries and among source units allows us to tentatively assess sediment yields from major subcatchments. Extreme yields and erosion rates are calculated for both the Karakorum Belt (up to 12,500 +/- 4700 t/km(2) year and 4.5 +/- 1.7 mm/year for the Braldu catchment) and Nanga Parbat Massif (8100 +/- 3500 t/km(2) year and 3.0 +/- 1.3 mm/year). These values approach denudation rates currently estimated for South Karakorum and Nanga Parbat crustal-scale antiforms, and highlight the major influence that rapid tectonic uplift and focused glacial and fluvial erosion of young metamorphic massifs around the Western Himalayan Syntaxis have on sediment budgets of the Indus system. Detailed information on bulk petrography and heavy minerals of modem Indus sands not only represents an effective independent method to constrain denudation rates obtained from temperature-time histories of exposed bedrock, but also provides an actualistic reference for collision-orogen provenance, and gives us a key to interpreting provenance and paleodrainage changes recorded by clastic wedges deposited in the Himalayan foreland basin and Arabian Sea during the Cenozoic. (C) 2004 Elsevier B.V. All rights reserved.