VALLEY EVOLUTION, UPLIFT, VOLCANISM, AND RELATED HAZARDS IN THE CENTRAL OF SOUTHERN

Jean-Claude THOURET(l), Gerhard WORNER(2), A. FINIZOLA(l), andA. LEGELEY-PADOVANI(3)

(1) Laboratoire Magmas et Volcans, Universite Blaise Pascal et CNRS, 63038 Clermont-Ferrand cedex, France ([email protected]) (2) Geoschemisches Institut, Goldschmidtstrasse I, Universitat Gottingen, 37077, Gottingen, Germany ([email protected]) (3) IRD, Centre lle de France, 34 rue Henri-Varagnat, 93193 Bondy cedex, France (A.Legeley­ [email protected])

KEY WORDS: Andes, Peru, valley history, volcanoes, , uplift, canyons, downcutting, hazards.

INTRODUCTION

Interpretation of satellite images, field work, and geochronology identify three principal types of volcanoes in the areas of Rio Cotahuasi-Rio Ocofia, and Arequipa in southern Peru (Figs. 1-3): (I) relatively youthful stratovolcanoes of to Recent age e.g. El (:S0.8 Ma) and older, larger complexes such as Nevado , (2) deeply eroded stratovolcanoes of Plio-Quatemary age (4-1.5 Ma) such as Nevado Solimana and , and; (3) subdued 'shield' volcanoes of Late age. While the former are rather varied in petrography and hence chemical composiotion oferupted magmas, the latter tend to be comprised more of monotonous mafic . Ignimbrites form of several hundreds of km2 in area and several tens of krrr' in volume are also observed and have been dated at middle Miocene (13-14 Ma), (5-2 Ma), and Pleistocene age (~I Ma) (Thouret et aI., 2001).

In the area of Arequipa, 13-14 M ignimbrites outcrop above the Jurassic in the Rio Chili valley that cuts the flank of the Western Cordillera (Fig. 2). The upper flanks of Rio Chili valley are built in part by flows and volcaniclastic sediments of Plio- age, whose sources are the Chachani massif and El Misti. Downcutting by 250 m has been achieved within the past 3 My (Fig. 2).

In contrast, the headvaJleys of the deepest canyons on Earth of Rios Ocofia, Cotahuasi and Colca NW of Arequipa, have been cut 2 km down in Miocene volcanic rocks and I km further down in Cretaceous intrusive and Jurassic sedimentary rocks. The headvalleys were repeatedly filled by pyroclastics and lava flows of Neogene age « I Ma, Fig. 3) and have subsequenty been recut below their original thalweg. These canyons are thus much deeper and older (middle Miocene) than Rio Chili.

641 II tS Slll

Fig. 1 showing the area of the canyons of Rio Ocofia and Rio Cotahuasi in the Western Cordillera, ignimbrites, and three types ofvolcanoes. Main geologic units and faults are also shown.

642 Marine sediments and conglomerates of Eogene age are uplifted to as much as 2000 m asl, in the area of Caraveli and on the east side of the Rio Ocofia (Fig. 3). The main palaeosurface, which has been mantled by the Huaylillas ignimbrites of lower Miocene age, is tilted and ist eastern parts uplifted from about 2500 m to 4000 m asl. in the area of Chuquibamba (Fig. 3). More than half of the uplift (about 2000 m) of the Western Cordillera postdates the emplacement ofthe Huayli11as ignimbrites (Kennan, 1999). This uplift, combined with the increase ofwater discharge from glacial sources since Upper Pliocene times triggered the downcutting by at least 1500 m of the Ocofia, Cotahuasi and Colca Canyons since the middle Miocene to Pliocene. Further downcutting of the canyon occured after the emplacement and infilling ofthe Pliocene lava flows and ignimbrites

EL MISTI NE

5600 m Polygenic 4800 puna surface of sw Miocene age lava flows of middle Pleistocene age 000 -5 ~a·old ran up if s the coastat c rz: 3200 m 3200 m a:UJ .. j Rio Vilor ·Pampa· surface cuts late PJiocene ignimbrite ---- and Rio Chifi Top 01ignimbriles 01 --- and eogene 2400 middle Miocene age + + t + .j. j conglomerates ! + ...... ,+_..:!:+--!:~~~~§~~ Coastal Ba/hafit Rio~~ 1200

120km 100 km 50km 10 km

Fig. 2 showing the Rio Chili valley through the Western Cordillera and the depression of Arequipa. El Misti stratcvolcano and three groups 01 ignimbrites (middle Miocene, Pliocene. and early Ptelstocene). The amount of uplift and downcutling is estimated using the Miocene surface remnants and the base 01ignimbrites and lava flows.

NNE N subdued volcanoes of upper Miocene age \ ignimbrite of Nevad~, ~~r~~MANA

Landslide Cerro Saracotto Pliocejneage Pleistocene eooom soocm and glacial Mafic shield of cirque " , .... Plio-Qusternary age Rear of lower Miocene 5200 ~ "- [ ...... ----,~ q.cs>' Iforeland basin 3600 .~ '<::: ~ .Uj igrr ---- -y - ~ Z~~~~~: uplift and deformed Jl B ~ g- _ ~ ~rnbrites 01 MIOfene age I S1 surface of lower Miocene ~ §~o :§-- & --~ _ __ ~rnnants , uplifted Moquegua formation of 7(~~8 l·~ ~; (at Pampa Cerro Cuno cun~ligocene to lower Miocene age 2800 2800 • .. 18-14M;i"?] ~ r~~nrs Rio Huarcayo \ t r Q :2_ La vesere-- ...... Pampa detlnti Piemonl 2000 2000 ~ C'(!>~ ~ ,:~~~ cc (at 10'7Mi'?) Pampa cerrc Cabo --:Ri'ii'O"""C;:':Ola"'h-ua-s,-·-\ '-0 ~ ~ ....§-lLaL BJanca El Alto I~ break S-ZMa?) 8 In slope Confluence of Rio Oeona t 400 end Rio CotahU8Si downcuttlng since Pllocene '90Jun ,5Dbn l00bn sokm 10km Fig.3 showing the canyons of the Rio Ocona and Cctahuasi through the western Cordtltera, three groups of volcanoes (upper Miocene, PJio-Ouaternary, and Pleistocene in age), and three groups of ignimbrites (Alpabamba, Huaylillas. and Sencca) . The amount of uplift and downcuding is estimated using the S1. 82. and 83 surface remnants.

CONCLUSIONS Digital elevation model (DEM) based on six digitized topographic maps (1:250.000) will be used to illustrate the relationships between the generations of ignimbrites and volcanoes, the uplift of the surface remnants and the down cutting of the canyons. The rapid rate of uplift of the western Cordillera and the distinct rate of the down cutting in the deep Rio Ocofia canyon versus that of Rio Chili are inferred from calculations using the DEM and Ar-Ar geochronology on volcanic rocks.

Downcutting of the deepest canyons on Earth has triggered huge landslides. Collapses involved the Miocene ignimbrites that form the highest canyons walls, e.g., above the town of Cotahuasi, and the edge of the

643 Huaylillas plateau near Chuquibamba (Fig. 3). Downcutting also triggered flank failures on the Plio-Quaternary volcanoes, such as the southwest flank of Nevado Solimana. Subsequent debris avalanches have dammed the upper course of Rios Cotahuasi and Rio Huarcaya in Pleistocene time. Downcutting continues and valley flanc and volcano instabilites pose a major hazard in these ultr-deep canyons. Further hazards are also related to potential dam breakouts that may trigger devastating debris flows toward the populated lower valleys.

REFERENCES

Kennan L., 1999. Large-scale geomorphology in the Andes: interelationships of tectonics, magmatism, and climate. In: M.A. Summerfield, ed., Geomorphology and Global Tectonics, p. 167-199, J. Wiley, New York.

Thouret J.-c., Finizola A., Fornari M., Suni J., Legeley-Padovani A., and Frechen M., 200 I. Geology of El Misti volcano nearby the city ofArequipa, Peru. Geological Society ofAmerica Bulletin, vo!. 113, n012, 1593-1610.

644 'IVI!RSI IH PA Ul. Geodynamique andine SABAT IER lnstitut de recherche pourle developpernent Andean Geodynarnics l DlJl.OlJSI III Geodinarnica Andina

Resumes etendus Extended abstracts Resumenes expandidos

Organisateurs Organizers Organizadores

Institut de recherche Universite Paul Sabatier pourle developpement Toulouse Paris France 2002

UNIVERSITE PAUL

lnstitut de recherche pOUf le de\leloppement •TOUI.OUSI< III

GEODYNAMIQUE ANDINE ANDEAN GEODYNAMICS GEODINAMICA ANDINA

5th International Symposium on Andean Geodynamics Universite Paul Sabatier, Toulouse, France, 16-18 Septembre 2002

Resumes etendus Extended abstracts Resumenes ampliados

Organisateurs / Organizers / Organizadores Institut de recherche pour le developpement Universite Paul Sabatier

IRD INSTITUT DE RECHERCHE POUR LE DEVELOPPEMENT Paris, 2002 COMITE D'ORGANISATION COMITE ORGANIZADOR ORGANIZING COMITTEE

P.Baby (IRD-Toulouse), I Darrozes (Univ. Paul Sabatier-Toulouse), I Deramond (Univ. Paul Sabatier-Toulouse), B. Dupre (CNRS-Toulouse), I-L. Guyot (IRD-Toulouse), G. Herail (IRD-Toulouse), E. Jaillard (IRD-Quito), A. Lavenu (IRD-Toulouse), H. Miller (Univ. Munchen), T. Monfret (IRD-Geoscience Azur), G. Warner (Univ. Gottingen)

Comite scientifique et representants nationaux Comite Cientitico y Representantes Nacionales Scientific Advisory Board and National Representatives

R Armijo (IPG, Paris), I-P. Avouac (CEA, Paris), R Charrier (Univ. Chile, Santiago), I-Y. CoBot (IRD, Geoscience Azur), L. Dorbath (IRD, Strasbourg), S. Flint (Univ. Liverpool), B. France-Lanord (CNRS, Nancy), L. Fontbote (Univ. Geneve), Y. Gaudemer (Univ. Paris VII), R Gaupp (Univ. Jena), F. Herve (Univ. Chile, Santiago), T.E. Jordan (INSTOC, Cornell), I Mojica (Univ. Bogota), O. Oncken (Univ. Potsdam), L. Ortlieb (IRD, Bondy), RI Pankhurst (Brit. Antartic Surv.), V. Ramos (Univ. Buenos Aires), P.Ribstein (IRD, Paris), C. Robin (Univ. Clermont-Ferrand), S. Rosas (Univ. Lima), F. Sabat (Univ. Barcelona), M. Schmitz (FUNVISIS, Caracas), R Suarez Soruco (YPBF, La Paz), M. Rivadeneira (Petroproducci6n), W.Winkler (ETH, Zurich).

APPUIS FINANCIERS FUNDINGS APPOYO FINANCIERO L'organisation de l'ISAG 2002 et les bourses accordees aun certain nombre de collegues latino-americains ont ete possibles grace au soutien financier de l'IRD (notamment de la Delegation al'Information et ala Communication), de la region Midi-Pyrenees, de l'Universite Paul Sabatier et de l'Andean Comittee de l'ILP.

© IRD,2002