J ournal oJ Glaciolog)" f'ol. 41, No . 137, 1995

Recession of Yanatnarey Glacier in Cordillera Blanca, Peru, during the 20th century

STEFAN HAST ENRATH

Department of Atmospheric alld Oceanic Sciences, Univasil» oJ W isconsill Jladison, 1225 West Day ton Street, Madison, r Visconsil! 53706, U.S.A .

ALCIDES AMES

Unidad de Glaciologia e Hidrologia, Electroperu, Huaraz, Peru

ABSTRACT. f or Yanamarey Glacier in Cordill era Blanca, P eru , mos tl y situ a ted between about 5000 and 4600 m , ma ps of the surface topogra ph y a t a scale of I : 5000 obtained by terres tri a l tri a ng ula ti on for 19 73, 1982 and 1988 and by ae ria l photogrammetry for 1948 a nd 1962 are com pa red with the glacier bounda ri es from a 1939 ma p a nd an und a ted m aximum ex ten t inferred from mo ra ine morphology. The g lacier leng th d ecreased from th e maximum , 2800 m to 1600 m in 1948, and to 1250 m in 1988, with a n accompanying d ecrease in a rca fro m 17 x 105 to 10 x 105 and th ence to 8x 105 m 2 The shrinkage of ice volume was 35x 106 m 3 fro m the max imum to 1948, and 29 x 106 m3 from 1948 to 1988, compa red to a total rem aining ice volume of a bout 25 x 106 m3 in 1988. This qu antita ti ve assess ment of m ass-loss ra les crea tes th e obse rvational basis for sensitivity studies of the clima tic forcing.

INTRODUCTION

Glaciers in the high mounta ins of the tro pi cs merit a ttention in the contex t of global change. M onitoring and doc um enta tion efIorts a t th e intern a ti ona l level (H ae berli and H oelzle, 1993; H ae berli and o thers, 1993 ) bea r out th e d earth of info rma ti on es pec ia ll y for the tropica l half o f the Earth. Successi\'e reconstructi ons have, however, affo rded a genera l overvie\\' of sec ula r glac ier vari a tions in the low la titudes . Thus, in E as t Afri ca, glaciers bega n to recede a fter 1880, as e\·id enced by a combina ti on of fi eld obsen 'a ti ons, expedition reports a nd numeri ca l mod eling (Has tenra th , 1984; Kruss, 1984). For Ne\\' Guinea (H ope and oth ers, 1976), Allison a nd Krus (1977 ) ded uced from numerical modeling a n onse t of glacier recession a round 1850. For the Ecuadori a n Andes, evalua tion of historical sources (H as tenra th , 1981 ) indi­ca tes a gradu al rise of the ice-equilibrium line since a t least the middle of the 19th century. For the mos t recent decades, con tin uing deCt-ease of the ice-covered area has been reported in the Venezuela n Andes (Schubert, 1992), and accelera ted \'olume loss has been measured on the largest outl et glacier of the Quelceaya lee Cap of southern Peru (Brecher and Thompson, 1993 ).

I 970s and 1980s includ ed repeated terres tri al ma pping of sur face topogra ph y a nd ice ex tent ofY anamarey Glacier in the Cordill era Bla nca . For 1962 a nd 1948 it proved

\vith this b ac kground , th e Cordill era Blanca of northern P eru (Fig. I ) m e ri ts pa rti cul a r a ttention, because it was the object of a systema tic glacier inventory and mapping in the 1930s (Kinzl, 1942, 1949, 1964) and of later work by the Unidad d e Glaciologia e Hidrologia, Elec troperu (Ames and others, 1988; K aser and others, 1990) . Glaciological monito ring by Electroperu in the

9' 30' S

o 10 km 20

77' 30'W

Fig. 1. Orientation map oJ CordiLLera Blallca with height c0 l1 t01lrS oJ 3000 and 5000 m; scale 1 : 1000000 . Rectallgle in southeast corner indicates area oJ ranamarey Glacier ill Figure 2. Insert in !ljJ/Jer-right corner shows locatioll oJ CordiLLera Blanca within Peru .

191

J ournal of Glaciology

possible to map the surface topography of this glacier from historical a ir ph otogr a phy. In addition , ice bound aries could be traced for the 1930s from an expedition acco unt (Kinzl, 1942, 1964), a nd for a maximum exten t from large m oraines . In the present paper, this diverse evidence is compiled and evalua ted to reconstruct th e length , a rea and volume ch a nges of Yanamarey Glacier over much of the pas t century. This is motiva ted by the need for qua ntita tive informa tion for ana lysis 0 [" the climatic fo rcing.

2. THE MAPPINGS

Information of varied detail and quali ty is availa ble for seven dates : a n unda ted maximum extent indica ted by moraines, and M ay-August 1939, 31 August 1948, 17 June 1962, 10 M ay 1973, 19 August 1982 and 10 August 1988 . The scale used in all m a ps is 1 : 5000.

The maps for 1973, 1982 a nd 1988, created by A. Ames, were produced by double-theodolite tri a ngula tion of a target rod held by a m obi le colla bora tor, from precision-surveyed control points on rock outsid e the glacier (Fig. 2) . These reference points are tied into the UTM coordin a te system (Instituto Geografico Militar, 1970, 1976) . Contour spacing was 25 m. From the ro ughness of the glacier surface, tolerance is es timated a t about 10 cm , although the surveying method yields a much higher precision.

8932.0 N

, /

/ /

/

/

250.0 E

/

/

The maps for 1962 and 1948 were produced by J. Sch erz, U niversity of Wisconsin- M adison, U .S.A. by stereo-photogrammetri c analysis of historical ver tical air pho tographs. The p articulars for 17 June 1962 are AF -60-17, M 134, frames 15420 and 1542 1, indica ted focal length 153 .035 mm. The d etails for 31 August 1948 are 45 7-88-B, fr a mes 2524-866 a nd 2524-867, with the same focal length assumed. A compila tion scale of I: 781 2 and subsequent en largement to scale I : 5000 was used with a contour interval of 25 m . T olerance is es tima ted at about 2 m . As r eference, four control poin ts (Fig. 2) were available, d e termined in the coordinate sys tem mentioned above. The 1962 map was compiled first, and supported by this the work for 1948 was undertaken . Coverage preclud ed m a pping of the upper reaches of the glacier in 1948.

A major accomplishment of the mountaineering a nd scientific expeditions of the Deutsch- Osterreichischer Alp enve rein to the Cordill era Bla nca in the 1930s (Kinzl, 1942, 1949, 1964) was a first inventory and mapping of the glaciers. A result of this effort is a topographic map a t scale I : 100000 based on terres tria l photogrammetry during May- Aug ust 1939. It proved possible to enlarge a pa rt of this ma p to the 1 : 25 000 scale of the official m a p sheet (Instituto Geografi co Milita r, 1970), in order to id entify prominen t terrain fea tures, and on this basis to transfer the 1939 boundaries of Yanam arey G lacier into the UTM coordinate sys tem, with a tolera nce of a bout 20 m. The 1939 ice ex tent thus obtained is a lso shown in Figures 2 a nd 3.

o

251.0 E

I m 500

+ 8931.0 N

Fig. 2. Map showing the ice extent of YanamaTey Glacier to the maximum inferredfTom large moraines andfor the da tes Ma)I-August 1939, 31 August 1948, 19 J une 1962, la M ay 1973, 19 August 1982 and 10 August 1988. Triangles indicate terrain-control points, and open circles the coordinated riference points llsed in the 1962 and 1948 maps. Contours at 100 m intervals are givenfor the 1988 status. 1988 extent of lake is shown by dotted line. Also entered is central longitudinal line with marks at 100 m spacing from up/JeT edge of glacier ( Fig. 3) . Scale 1: 20000.

192

Hastenrath and Ames: Recession of Yanamarry Glacier during the 20th centlll)!

2500 m 2000 I

1500 1000 I

500 I

o I I I ' , I , , , I' Id 'I' I '

5100 ··· ··· ···· MAX ----- 39

73 82

--88 A SURFACE TOPOGRAPHY

--48 ----- 62 ,<'" BEDROCK

.. ' m

62 4500 .. ' . .......

MIX 39 48 173 82 88 , I' IH 'I ' I, I

1500 B WIDTH

2500 m 2000 1500 1000 500

500

o I o

Fig. 3. LongitudinalproJiles along central line shown in Figure 2; scale 1: 20000. a. Longitudinal-vertical cross-section with no vertical exaggeration . Cross-hatching denotes bedrock; dotted, broken, solid, dash-dotted, dotted , broken and solid Lines denote the ice sUlface topograph),for the nWJ..imunI ice extent and 1939, 1948, 1962, 1973, 1982 and 1988, respectivef), . Vertical tick marks show terminus /Jositiol1s, illcLuding those for 1939 and the maximum extent. b. Glacier width plotted against distance from head, as measured along the Lines passing through the equidista7lt points pe/pelldiwlar to the central longitudinal Line shown in Figure 1,for the ma rimlllll extent and 1939, 19-18, 1962, 1973, 1982 and 1988, ill m. Tick marks and labels 011 the middle horizolltal scale indicate the terminus /Jositions for these dales .

Fina ll y, a maximum ice ex tent, without date, was determ ined from a sys tem of la rge mora ines that a re prominen t on air photographs ( 1962 63, AF -60-17: M 132, 19 June 1962, fram es 16701 - 16702; M1 63, 12 Jul y 1962, frames 2 1094-2 1095; M 337, 23 1\1ay 1963, fi'ames 4 1071 -4 1072). Using identifiable terrain fea ture fo r reference, these \\'ere transferred on to the existing I : 25 000 shee t (lnstituto Geografico 1\I ilitar, 1970) and thus into the UTM coordin ate system, with a tolerance of a bout 20 m. The maxim um extent of th e lower part of Ya namarey Glacier thus obtained is a lso shown in Figures 2 5.

3. EVALUATION

The diverse evidence described in sec ti on 2 n eed ed to be evaluated in a n interna lly consistent contex t . T o that end , the maps fo r the various dates were brought into the same sys tem of coordinates of (he nationa l ma p as indica ted above. Th e entire informa tion cO l1lent of each of the 1988, 1982, a nd 1973 maps, namely coordin a te sys tem, control point , glacier bou ndary, and height contours as well as centra l longitud in a l li ne with points at lOOm spacing (Figs 2 and 3), was digita ll y transferred to a data base. Th e same was done for the 1962 a nd 1948 maps. For the epochs 1939 and " maximum ex tent" the input consisted of the ice bound ary in th e lower pa rt of

the g lacier, togethe ,- with the een tral line and eq uidistant po ints. Th e subsequent work co uld proceed digitally .

From th e height contours, the surface topography was linea rl y interpola ted on to a rectangular grid with 2m spacing. Forming differences a t each grid point gave the surface lowering between sun'eys, or the ice-thickness change.

The glacier was divided into lOOm wide bands, bo unded by lines passing through th e aforementioned equidistant points perpendicul a r to the centra l long­itudina l line (Figs 2 and 3). Areas a nd a rea ch a n ges of th e. e bands were eva luated digi ta lly. Also, for these dom ains, th e vol ume changes between surveys were determined by in tegrating the ice-th ickness cha nge over the respec ti ve areas.

Estima tes o f ice thickness, a nd thus volume, were o bta ined for the 1982 date. Essenti a l fo r this was the bed rock topograph y a long the centra l longitudinal line (Fig. 3a). This was availab le with a tolerance of about 10 m from numeri cal modcling to be d etailed elsewhere in due course. Assuming a parabo lic shape of th e bedrock, th e center-Iine d cp th a nd the obse rved width d efine the bedrock topograph y along the lines placed a( lOO m in tervals perpendicu lar to the cen (ral line. These values were digitized a t 50 m spac ing a nd thf: 11 interpola ted on to a rec tangular g rid wi th 2 m cell size as described above. Thickness cha nges for the 100 m wide bands were th en eva lua ted digita lly.

193

J ournal of Glaeiology

For the 1939 d a te and [or the maxImum ex ten t, ice thickn ess was es timated from the available topographic shee t (Instituto G eograIico Milita r , 1970) as the difference in elevation between the center line and the ice m a rgin. Estimates were m ade at equidistant center-line points d own-glacier from the 1500 m longitudinal position. For the maximum position, the m a rgin is marked by the la tera l moraines. I ce thickn ess and volume [or th e 100 m wide bands were then evalua ted as described a bove.

4. CHANGES IN ICE EXTENT AND VOLUME

The broadly southwestward-facing Yanamarey G lacier extends from its steep upper rea ches above 5000 m to a terminus now well a bove 4600 m . Since the earl y 1970s the ice front has started to calve in to a sm a ll lake contained behind an arc of m orain es a nd a rock threshold. The glacier is now a bout 1 km long, with an a rea of less than 1 km2 An impression of the topogr aphy, from the mounta in peaks behind to the valley below the glacier, is offered by Figure 3a. A complex of large moraines is taken as marking the " maximum ex tent" o[ the glacier, a lthough other la rge moraine sys tems are found ye t further down th e valley as witnesses of earli er glacia l events; these, however, will not be of concern here.

From a length of 2800 m a t the maximum, the glacier had retreated to 1850 m by 193 9, with progressive further shrinkage in length and area during the subsequent intervals (Fig. 2) . The changing ice-surface topogra phy is illustra ted in Fig ure 3a. From " m aximum " to 1948, and again from 1948 to 1988, th e surface lowering exceeded 70 m in the lower part of the glacier . Accompan ying the re treat of the glacier to ngue from the " maximum" to 1939 was a m assive decrease in the width and a rea (Fig. 2) in the lower p a rt of the glacie r (Fig. 3b). After 1948, th e glacier was contained in a topographically more constrained d om ain, and accordingly the dras tic thickness changes (Fig. 3a) were accompa nied by compa ra tively moderate decreases in the width and area, prima rily in the lower part of the glacier (Fig . 3b).

The longitudinal pattern of changes in area, thi ckn ess a nd volume be tween the survey d a tes is depic ted in Figure 4a- c. For all three elem ents the decreases were m os t dras ti c n ear the progressively receding terminus. Thus, from " m aximum" to 1939 and thence to 1948 a rea decreases were la rgest a t longitudinal positions down­glacier from 1300 m, while during the subsequent time in tervals the ch anges were concentrated between 1600 and 1300 m. D ecreases in thickness a nd volume ar e found to be larges t som ewhat further up-glacier, a t longitudinal distances of 3000- 1500 m during " maximum" to 1948, a nd a t 1300- 700 m from 1948 to 1988 (Fig. 4b and c). Paralleling the progressive retreat of the glacier tongue, the la rgest thickness and volume changes tend to occur further up-glacier in later time intervals. Ch a nges are generally small a t longitudinal distances less tha n 500 m.

Figure 5 presents the secula r variations [or the glacier as a whole. The ice volume (F ig. Sa) decreased by 35 x 106 m3 from " maximum " to 1948; by 22 x 106 m3

from 1948 to 1982; and by 7 x 106 m3 from 1982 to 1988. From the inform ation give n in section 2, the error tolerances corresponding to these three time intervals

194

1940 50 60 70 80 90

f-----T-+-I , ,I , , I I , , 1 1 1 I, , I

............ A VOLUME

+50 (reference 1982)

,06m3

0

-20 MAX 39 48 62 73 82 88

f-----\--+-I , 11 11 11 , 1 I , 11 11 +2 B AREA

106m2 "--..

+1

0 MAX 39 48 62 73 82 88

f-----\--+-I 11 , 11 11 , 1 1 11 1 1 1

C TERMINUS ELEVATION

4600

m

4500

4400 M~ 39 48 62 73 82 88

f-----\--+-I I I 11 11 11 1 I , 11 1 1 3000 o LENGTH

m '" 2000

1000 ' I

1940 50 60 70 80 90

Fig. 5. Secular variation of Yanamarey Glacier. a. volume, measured relative to 1982 volume, which is estimated to be 32 x 106 m3 b. area. c. terminus elevation . d. length.

are estimated to decrease £i'om 20 x 106 to 2 x 106 to 0.2 X 106 m3

. From m odeling, referred to in section 3 and to be detailed elsewh ere in due course, the absolute ice volume of the entire glacier in 1982 is estimated to be 32 x 106 m3

, with a tolerance of about 5- 10 x 106 m3. The

tota l a rea (Fig. 5b) is estimated to be 17 x 105 m2 for the maximum ex tent, compared to 8 x 105 m 2 in 1988. The terminus elevation (Fig. 5c) changed by a large amount from m aximum to 1939, but then a t a more rapid rate as a steep bedrock threshold was being p assed. The leng th (Fig . 5d ) also decreased substanti a ll y from the maximum to the 1988 conditions.

5. CONCLUDING REMARKS

As indicators of clim a ti c change, vari a tions of tropical glaciers are relevant in the global contex t. In this spirit, changes in the length and area of glaciers worldwide a re now being regularly compiled and published in an inter­na tional effort (Haeberli and Hoelzle, 1993). Although assessm ent of mass ch anges is needed for the stud y of the

Hastenrath and Ames: Recession of ranamarey Glacier during the 20th centw)I

50

103 m2

o

100 m

o

6 106 m3

4

2

o

A ~AREA

2

B ~ ICE THICKNESS

C ~ VOLUME

MAX-39 39-48 48-62 62-73 73-82 82-88

o

o

2500 m 2000 1500 1000 500 o Fig. 4. D ecreases over the time intervals maximllln- 1939, 1939-48, 1948- 62, 1962- 73, 1973 82 and 1982- 88, jor 100 m wide bands as defined il1 Figures 2 and 3. H orizontal scale 1 : 40000. a. Area. b. ice thickness. c. volume.

energy processes involved, even information on leng th and area changes is useful. Thus, ass uming a priori a n increase of air tempera tu re as the sole cause of glac ier retrea t, O eri emans (1994) used this d a ta source to compute a co rresponding a tmospheric-warming rate. J 11 a similar vein , Thompson and others (1993 ) interpreted ice-co re records in terms of tempera ture cha nge without considera­tion of ice-mass and energy processes. Tha t forcings other than a tmospheri c warming may be instrumenta l is shown by numeri cal modeling of th e ice d ynamics, hea t-budget sensiti vity studies and diagnos ti cs of secul ar circula tion changes for the glaciers of Eas t Africa (Kruss, 1984; H astenra th and Kruss, 1992; H as tenra th , 1994) .

vVith this m otivation, Ya na m a rey Glacier was a ta rge t of opportunity fo r the assessm en t of volume cha nge ove r

much of this century. This is impor tant because o nl y cha nges in mass can properly be conver ted into energy equiva lents in phase-change processes. Such informa tion is ra re especiall y fo r the tropical ha lf of the Earth . The qu a ntita tive reconstruction of ice-mass changes presented her e for Yanama rcy Glacier se rves as a basis for ongoing sen sitivity analyses a imed at inferring th e clim ati c fo rcing of the obse rved glacier shrinkage .

ACKNOWLEDGEMENTS

This stud y was supported by U .S. N a tional Science Fo und a tion grant N o . EAR-92 17211. At the University of vVisconsin, J. Sch erz performed the stereo-compila tion

195

J ournal oJ Glaciology

from the 1962 and 1948 a ir photography, a nd 1. Greischa r and D . Polzin assisted with the da ta processing and graphics. Comments by R . H ooke, 1. Thompson and an anonymous reviewer on an earlier version of this paper a re gra tefu lly acknowledged.

REFERENCES

AlIi son, I. and P . Kruss. 1977 . Esti ma tion of recent cl imate change in l rian J aya by numerical modeling o[ its tropical glaciers. Arc/. Alp. Res., 9( 1), 49- 60.

Ames, A. and 7 olhers. 1988. Inuenlario de glaciares del Peni. Hua raz, U nidad de Glaciologia e Hidrologia.

Brecher, H. H . a nd L. G. Thompson. 1993. M easuremen t of the retreat of Qori K alis glacier in the tropi ca l Andes of Peru by terrestrial phorogrammetry. Phologramm. Eng. Remole Sensing, 59 (6), 10 17 1022.

Haeberli , W. and M . Hoelzle, comps. 1993 . Fluclualions oJ glaciers 1985 90 ( Vol. 6). W allingford , l nterna tion a l Association of H ydrologica l Sciences; Nairo bi , U nited Na tions Environment Progra m me; Paris, UNESCO .

H aeberli , W ., E. H en 'en and M . H oelzle, eds. 1993. Glacier mass balance bul/elin . Bul/elin No. 2 (1990-1991) . Walli ngford , 1 nterna tional Association of H yd rological Sciences; Nai robi, U n ited Na tions Environmen t Programme; Paris, U N ESCO.

Hastenra th, S. 198 1. The glacialion if Ihe Ecuadorian Andes. R ot terdam, Balkema .

Hastenra th , S. 1984. The glaciers oJ equalorial Easl Africa. Dordrecht. etc .. D. Reidel Publishing Co.

Hastenra th , S. 1994. R ecession of tropica l glaciers. Science, 265, 1790 179 1. H as tenra th , S. and P . D . Kruss. 1992. The d ra matic retreat of J..101ln t

K enya's glaciers be tween 1963 and 1987 : greenhouse forcing. Alln. Glaciol., 16, 127 133.

Hope, G. , U . Radok, J. A . Peterson and 1. Allison. 1976. The equalorial glaciers oJ New Guinea: resulls oJ Ihe 1971 1973 AlIslraliall Vllil/ersilies' eX/ledilions 10 h iCln Jaya. R ot terdam, Balkema.

Ins tituto Geogra fi co Milita r. 1970. Carla naciollal 1:25,000. Hacienda Breque, hojCl 20i-/ I1-SE. Lima, Instituto Geogra fi co Mili ta r.

I nsti t ll to Geogra fi co Mili ta r. 1976. Carla nacional 1: 100,000. Recuay, hoja 20i. Lima, InstilUto Geogr;i fi co Mi li ta r.

K aser , G., A. Ames a nd M . Zamora. 1990 . G lac ier nuctua tions and clima te in the Cordillera Blanca, Peru . AnlZ. Glaciol., 14, 136 140.

K inz l, H . 1942. G letsch erkundli che Beglei tworte zu r K artc der Corclill era Blanca (Peru ). Z. Glelscherkd., 28 (1/2), 1 19.

Kinzl, H. 1949. Die Verg letscherung in d er S lidh ii lfte der Cord illera Blanca (Peru ). Z. Glelsdzerkd. GICl<.ialgeol. , 1 (1), I 28.

Kinzl, H . 1964. Begleitwort zur Karte l:l OO 000 der Cordillera Bla nca (Peru ). Slid teil. VIIiss . Alpenvereinsh. 1.

Kruss, P . 1984. Clima te cha nge in East Africa: a nu merica l simulation from the 100 yea rs of terminus record a t Lewis Glacier, J..;l o un t K enya. Z. Glelscherkd. Gla<.ialgeol., 19 (1), 1983,43- 60.

O erlemans, J. 1994. Quan tifying global warming fro m the retreat o f g laciers . Science, 264, 243- 245.

Schubert, C. 1992. T h e g laciers of the Sierra Nevada de M ericl a (V enez uela ): a pho tographic com parison of recent deglacia tion. Erdkunde, 46(1), 58-64.

T hompson. L. G. alld 6 olhers. 1993. "R ecent warmi ng": ice core evidence (ro m tropical ice cores wi th emphas is on Centra l As ia. Global alld Planelm), Change, 7 (1- 3) , 145 156.

MS received 7 MaTch 1994 and in TevisedJorm 7 J uly 1994

196