Recent trend of climate change in southern South America

When glacier variations in Patagonia are discussed from a climatic point of view, meteorological data at Puerto Aisen and Punta Arenas are often cited, from which we cannot identify any significant trends in air temperature and annual precipitation (Warren & Sugden, 1993). Although these stations provide relatively long, continuous records from around Patagonian glaciers, they are located far from glaciers, the former station being about 150 km to the north of the northern margin of the NPI and the latter being about 250 km to the south of the southern margin of the SPI.

Here, we select two minor meteorological stations near the SPI, namely the Station Islote Evangelistas on a small island about 150 km to the southwest of the southern end of the SPI and the

Figure 46.4 Annual thickness change in ablation areas of Upsala, Perito Moreno and Tyndall Glaciers (SPI), and Soler Glacier (NPI), which terminated on land before around 1992 and now calves into a newly formed lake. (Sources: Naruse et al., 1995; Skvarca & Naruse, 1997; Naruse & Skvarca, 2000; Skvarca et al., 2004).

Moreno

Soler

Tyndall

Upsala

»--------------a i—i—i—i—i—"i—i—i—i—i—i—i—r i i r 1984 1987 1990 1993 1996 1999

Year

TEMPERATURE - ISL. EVANGELISTAS (52.4 S, 75.1 W, 49 m)

TEMPERATURE - ISL. EVANGELISTAS (52.4 S, 75.1 W, 49 m)

a

m ■

■ • ■

m *

|i

■ ■

■ ■1

■ ■

*

«

■ ■

1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990

YEAR

1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990

YEAR

6000

5000

4000

ar

e

3000

E

(m

2000

1000

PRECIPITATION - EVANGELISTAS (52.4 S, 75.1 W, 49 m)

1880 1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990

YEAR

Figure 46.5 Annual mean air temperature and annual precipitation between 1900 and 1988 at the Station Islote Evangelistas, Chile (Rosenblüth et al., 1995). Reproduced by permission of the Bulletin of Glacier Research.

Station Lago Argentino at Calafate about 60 km from the terminus of Perito Moreno Glacier. Fluctuations in annual mean temperatures and annual precipitations at these two stations are shown in Figs 46.5 & 46.6. Although year-to-year variations are considerable, we can notice a slight rising trend in temperature during the 20th century or the past 50yr. Rosenbluth et al. (1995) stated that, south of 46°S a definite warming trend appears in the records, both on the Pacific and the Atlantic coast, especially in the past two decades. As to annual precipitations, weak decreasing trends are recognized at both stations.

LAGO ARGENTIO

3 9 15 21 27 33 39 45 51 57 63 69 75 81 87 6 12 18 24 30 36 42 48 54 60 66 72 78 84 90

YEARS

600 500

"E

JL 400 N IO

100 0

TL

«V

ii

Tir?

*SF< *

3 9 15 21 27 33 39 45 51 57 63 69 75 81 87 6 12 18 24 30 36 42 48 54 60 66 72 78 84 90

YEARS

3 9 15 21 27 33 39 45 51 57 63 69 75 81 87 6 12 18 24 30 36 42 48 54 60 66 72 78 84 90

YEARS

= ANNUAL

TENDENCY

MOVING AVERAGE

Figure 46.6 Annual mean air temperature and annual precipitation between 1937 and 1990 at the Station Lago Argentino, Argentina (Ibarzabal y Donangelo et al., 1996). Reproduced by permission of the Bulletin of Glacier Research.

An analysis based on the NCEP Reanalysis data (Climate Data Center, National Oceanic and Atmospheric Administration) indicated that air temperature at 850hPa over southern South America rose with a mean rate more than 0.05 Kyr-1 between 1948 and 1989 (K. Kubota, unpublished data). This result, as well as the trends of temperature and precipitation at the above two stations, seems to support the interpretation of recent recessions of most glaciers in Patagonia.

At Perito Moreno Glacier, based on the ablation measurements in the summer of 1993-1994 and a degree-day calculation, mean annual ablation rates over the past 30 yr were estimated as about 14 ± 2m in ice thickness (Naruse et al., 1997). Thus, a range between the maximum and the minimum of the annual ablation caused by year-to-year temperature variations can be considered as about 4 m in thickness, which is much smaller than the measured annual ice thinning. It was concluded that the change (rise) in temperature alone cannot explain the thinning rate of 11myr-1 at Upsala Glacier. Therefore, the recent recession of the glacier must be caused by a non-climatic effect, that is, the glacier dynamics.

46.5 Calving dynamics and glacier variations

Studies on calving dynamics of glaciers mostly have been made at fjords in Alaska and the Arctic. It was known that although some Alaskan tidewater glaciers were retreating catastrophically, others were oscillating or advancing slowly (Mercer, 1961). Change in the terminus position of a calving glacier can be expressed as dL/dt = U - Uc (1)

where L is the terminus position, Ui is ice velocity at the terminus and Uc is the calving rate (speed), which is defined by the volume rate of iceberg discharge (including the frontal melting) divided by the cross-sectional area of the terminus.

Calving rate is often represented empirically with a linear relation of water depth at the terminus. In Fig. 46.7, several empirical relations are shown. It is noted that calving rates of tidewater glaciers are more than five times larger than those of freshwater calving glaciers. This is due to the effects of tide, water current, salinity, and so on. These calving relations clearly demonstrate the typical characteristics of tidewater glaciers: when the terminus of a glacier retreats into deeper water from the moraine shoal or the bed rise, the rate of calving increases with the relation (Fig. 46.7), which may lead to the further recession of the glacier by Equation (1).

It is sometimes pointed out that back-stress (or back-pressure) arising from the shoal or islands is crucial to the dynamics or stability of a glacier. Meier & Post (1987) explained rapid disintegration of grounded tidewater glaciers owing to a feed-back process including back-pressure: that is, retreating^decrease in back-pressure^longitudinal stretching^ice thinning^decrease

0 100 200 300 400 500

0 100 200 300 400 500

Figure 46.7 Relationships between calving rate and water depth near the glacier termini. Dots and their regression line C are obtained from freshwater calving glaciers in Patagonia (Skvarca et al., 2002), the lines A and B are empirical relations for fresh water proposed by Funk & Rothlisberger (1989) and Warren et al. (1995), respectively, and the line D is derived for tidewater glaciers in Alaska (Brown et al., 1982).

Average water depth (m)

Figure 46.7 Relationships between calving rate and water depth near the glacier termini. Dots and their regression line C are obtained from freshwater calving glaciers in Patagonia (Skvarca et al., 2002), the lines A and B are empirical relations for fresh water proposed by Funk & Rothlisberger (1989) and Warren et al. (1995), respectively, and the line D is derived for tidewater glaciers in Alaska (Brown et al., 1982).

Figure 46.8 The frontal part of Upsala Glacier calving into the western arm of Lago Argentino, in November 1993 (upper photograph) and March 1999 (lower photograph), viewed from the control station of field surveys on the eastern (left-hand side) bank. Width of the glacier is about 3 km. The tributary glacier flowing from the western valley is Bertacchi Glacier, and the snow-covered, pyramidal peak is Cerro Cono (2440 m). (Both photographs taken by P. Skvarca.) (See www.blackwellpublishing.com/ knight for colour version.)

Figure 46.8 The frontal part of Upsala Glacier calving into the western arm of Lago Argentino, in November 1993 (upper photograph) and March 1999 (lower photograph), viewed from the control station of field surveys on the eastern (left-hand side) bank. Width of the glacier is about 3 km. The tributary glacier flowing from the western valley is Bertacchi Glacier, and the snow-covered, pyramidal peak is Cerro Cono (2440 m). (Both photographs taken by P. Skvarca.) (See www.blackwellpublishing.com/ knight for colour version.)

lo Islands above water cn cn cn cn

Figure 46.9 Longitudinal profile of the bedrock in the lower reaches of Upsala Glacier. Horizontal distance is measured northward from 50°S. Small islands are exposed to the south of the 2-km position in the western part of the lake. Glacier front in 1997 is approximate, and the fronts in 1998 and 2000 are located at about 6km and 6.7km, respectively (Naruse & Skvarca, 2000; Skvarca et al., 2002). Reproduced by permission of the International Glaciological Society and the Instituto Antárctica Argentina.

(North) Longitudinal Distance (km) (South)

Figure 46.9 Longitudinal profile of the bedrock in the lower reaches of Upsala Glacier. Horizontal distance is measured northward from 50°S. Small islands are exposed to the south of the 2-km position in the western part of the lake. Glacier front in 1997 is approximate, and the fronts in 1998 and 2000 are located at about 6km and 6.7km, respectively (Naruse & Skvarca, 2000; Skvarca et al., 2002). Reproduced by permission of the International Glaciological Society and the Instituto Antárctica Argentina.

in subglacial effective pressure^furtherstretching^calving^ further retreating.

By analysing data from Columbia Glacier, it was shown that, during the retreat, the thickness at the terminus appears to be linearly correlated with the water depth (Van der Veen, 1996). This is the 'height-above buoyancy model', which states that the position of the terminus is controlled by such geometric factors as ice thickness and water depth in order for the thickness in excess of flotation at the terminus to become about 50 m. However, the physical basis of this model has not been fully clarified.

Recent fluctuations of Upsala Glacier are worthy of note. From 1945 to 1978 the glacier had been almost stable, and the retreat started in 1978 (Fig. 46.3). A remarkable change in the front part of the glacier from 1993 to 1999 clearly can be seen in the photographs (Fig. 46.8). The width-averaged front positions of Upsala Glacier are indicated in Fig. 46.9. The front retreated remarkably in the periods of 1981-1984, 1990-1994 and 1996-1998.

Measurements of ice-thickness with ice radar have not been conducted successfully at this glacier owing to heavy crevasses and a large amount of water within the glacier. Instead, water depth surveys carried out in 1997 at the proglacial lake made it possible to reveal the subglacial topography under the glacier terminus in the period from 1970 to 1996 (Fig. 46.9). Bedrock rises were found to spread out from the exposed islands in the western part of the lake. Naruse & Skvarca (2000) proposed the following model. Between 1978 and 1990, the western half of the glacier terminus was located at the bedrock rises, which suggests that the front fluctuations were strongly regulated by the basal drag. During 1990-1993, the glacier terminus shifted upstream from the bedrock rises, and then the glacier was stretched longitudinally (with a strain rate of 0.22 yr-1 deduced from the continuity consideration), which resulted in ice thinning. This process may cause extensive calving, so that the glacier retreats further.

At San Rafael Glacier, Warren (1993) stated that, on annual to decadal time-scales up to about three decades, glacier oscillations may not be related directly to climate change. When we consider the glacier fluctuations at the longer time-scales of more than several decades or a century, however, climatic effects become more dominant. Thus, fluctuations in the calving terminus of glaciers over a short time-scale, such as a few decades or so, are considered to be controlled by non-climatic effects, specifically topographic, dynamic and subglacial hydrological conditions near the glacier terminus.

Was this article helpful?

0 0

Post a comment