The Editor

Journal of Glaciology

Sir

Climate change in the southern Andes has been marked by a loss of approximately 100 m of ice-surface thickness from the 3200 km² Hielo Patagónico Norte (HPN; northern Patagonia ice field) over the last 120 years (Winchester and Harrison, 1996). Because of the remoteness and inaccessibility of the ice field, centring on 47°00′ S, 73°30′ W, much of the work investigating change has focused on the movements of glacier ice fronts. Ice-front positions have been extrapolated from aerial photographs (Aniya 1985, 1987, 1992; Wada and Aniya, 1995), scanthistorical sources, dendrochronology and lichenometry (Winchester and Harrison, 1996). Ice-surface velocity has been directly measured for only two ot the 30 or so glaciers that descend from the ice field, Glaciares San Rafael and Soler (f1).

Fig. 1. Hielo Patagónico Norte and its glaciers.

This note records an indirect estimate of a mid-valley ice-surface velocity on Glaciar Arenales obtained from aerial photographs taken in 1975 and 1979. In February 1996 we investigated Glaciares Colonia, Arco and Arenales on the southeastern side of the ice field. All three glaciers begin their descents from the ice field over icefalls; Glaciar Arenales then flows eastwards for around 8 km before joining Glaciar Colonia coming in from the northwest. Our velocity estimate was based on the distance travelled over a 4 year period by a mass-movement feature near the Arenales valley midpoint (f2). The 1:50 000 regional map (Lago Colonia 4715–7300, No. 381) drawn from the 1975 aerial photograph (see f2) shows a comma-shaped feature (at 47°15′ S, 73°18′30′′ W) some 750 m longand 150 m wide extending out from the southern valley wall to approximately one-third of the way across the Arenales surface and about 3.5 km above the Colonia junction. Thisfeature, which we identified on the ground as a large rock-fall/debris-flow complex (f3), with the comma form created by the differential flow rates of surface ice, also appears onthe 1979 photograph. Measurement of the distance covered by the tip of the feature between the two dates provides an approximate rate of surface movement over the 4 year period of 150 m year⁻¹ or 0.41 m d⁻¹.

Fig. 2. The Arenales, Colonia and Arco glacier valleys showing the rock-fall/debris-flow complex in its 1975and 1979 positions.

Previous Surveys

The ice-surface velocity of Glaciar San Rafael has been measured on three different occasions (Naruse, 1985; Harrisonand Winchester, 1992; Rignot and othes, 1996). All three surveys were in close agreement, showing ice-surface velocity near the terminus averaging 17–17.5 m d⁻¹ over the 12 year period during which the studies took place. Although agreement over the ice-front movement rate is interesting per se, the velocity of this tidewater calvingglacier is related partly to calving flux (Warren and others, 1995). and thus is likely to be unrepresentative of the velocity of most of the other HPN outlet glaciers.

The other direct measurement of ice-surface velocity was carried out on the 8 km long valley section of the land-based Glaciar Soler on the eastern side of HPN in 1985 (Naruse, 1987; Naruse and others, 1992). The survey showed the glacier’s surface moving with an average velocity, between 21 survey stations, of 0.41 m d⁻¹. At around its valley midpoint the rate was 0.49 m d⁻¹, with a rate of 0.39 m d⁻¹ at the point above this. An indirect approach to velocity measurement using aerial photographs was also applied to Glaciar Soler by Aniyaand Naruse (1987) who, by dividing a number of annually formed ogive bands by the length of the glacier section containing them, derived an average ice-surface velocity over a 32 year period of 170 m year⁻¹ (0.47 m d⁻¹).

Discussion

The extraordinary similarity in the surface velocities of Glaciares Soler and Arenales, averaging 0.39–0.49 m d⁻¹ around their midpoints, is perhaps surprising considering their differences. The Soler, according to Naruse (1987), has an average slope angle of 2.9° and width of 1.5 km. compared with the average 1° slope angle and 2 km width Arenales which, after joining Glaciar Colonia, flows to a joint terminus 3.5 km further downslope. Although it is difficult to be very precise when using aerial photographs with-out specialist equipment, the approximate accuracy of our indirect approach to measurement of ice-surface velocity is supported by the degree of correspondence between the indirect and direct values obtained by Aniya and Naruse (1987) over long and short time periods respectively.

Fig. 3. A partial view of the southern flank of Glaciar Arenales, with the ellipse of debris material to be seen (arrowed) on its surface in the middle distance, about 3.5 km from the camera. (Photograph by V.W., January 1996.)